When The Going Gets Tough… How Action Versus State ... dissertation.pdfWhen The Going Gets Tough… How Action Versus State Orientation Moderates The Impact Of Situational Demands

When The Going Gets Tough…

How Action Versus State Orientation Moderates The Impact Of Situational Demands

On Cognition, Affect, And Behavior

Nils Jostmann

Vrije Universiteit Amsterdam

VRIJE UNIVERSITEIT

When the going gets tough...

How action versus state orientation moderates the impact of situational demands on cognition, affect, and behavior

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus

prof.dr. L.M. Bouter, in het openbaar te verdedigen

ten overstaan van de promotiecommissie van de faculteit der Psychologie en Pedagogiek

op vrijdag 2 februari 2007 om 13.45 uur in de aula van de universiteit,

De Boelelaan 1105

door

Nils Bernhard Jostmann

geboren te Bielefeld, Duitsland

promotor: prof.dr. G.R. Semin

copromotor: dr. S.L. Koole

When The Going Gets Tough… 4

Contents

Chapter 1 General Introduction ................................................................................................. 5

Chapter 2 Cognitive Shielding................................................................................................. 23

Study 2.1.............................................................................................................................. 31 Study 2.2.............................................................................................................................. 36

Chapter 3 Affective Shielding ................................................................................................. 48

Study 3.1.............................................................................................................................. 55 Chapter 4 Behavioral Shielding ............................................................................................... 65

Study 4.1.............................................................................................................................. 75 Study 4.2.............................................................................................................................. 85 Study 4.3.............................................................................................................................. 93 Study 4.4.............................................................................................................................. 97

Chapter 5 General Discussion................................................................................................ 112

Endnotes................................................................................................................................. 126

Summary................................................................................................................................ 132

Samenvatting (Summary in Dutch) ....................................................................................... 135

Zusammenfassung (Summary in German) ............................................................................ 138

References.............................................................................................................................. 141


Chapter 1

General Introduction1.1

In their professional, educational, and social lives, people often encounter a multitude of

demanding commitments. At the workplace, people may have an impatient boss on their back to

ensure fast job completion while adhering to high product standards. Educational environments

often press students to manage complex working schedules under compelling time constraints.

People also have to live up to various social expectations such as visiting a sick family member,

organizing an important neighborhood meeting, or settling a dispute between two friends.

Managing such everyday demands is not without costs. Meeting high demands can drain people’s

energy, make them more forgetful and less accurate, and leave them sulky and frustrated. In severe

cases, high demands may cause stress and even serious health problems (Lazarus & Folkman, 1984;

Segerstrom & Miller, 2004). It is therefore important to learn how people can effectively cope with

demanding situations (Hockey, 1997; Schönpflug, 1983; Zeidner & Endler, 1996).

In the present dissertation, I examine the role of action control in coping with high demands.

Action control refers to the mental processes that allow people to form, maintain, and implement (or

disengage from) their intentions (Kuhl & Goschke, 1994a). Building on Personality Systems

Interactions (PSI) theory (Kuhl, 2000, 2001), I suggest that efficient action control allows people to

cope more effectively with demanding situations. Efficient action control skills should therefore

constitute a key moderator of how people cope with demanding situations. Individuals who are

more skilled at action control, or action-oriented individuals, can be expected to function more

optimally under high demands compared to individuals who are less skilled at action control, or

state-oriented individuals. Indeed, this dissertation will try to show that, under demanding

conditions, action-oriented individuals display better working memory, more efficient affect

regulation, and more efficient implementation of difficult intentions relative to state-oriented

individuals. In the remainder of this introductory chapter, I discuss a theoretical analysis of action


control and performance under high demands. Next, I provide an overview of the remaining

chapters of the dissertation, which report empirical tests of this analysis.

On the Regulation of Action

High demands usually emerge when people attempt to prepare or enact some form of action.

In order to understand the dynamics of high demands, it is therefore useful to first examine the

workings of human action. To a large extent, human action is based on routines and guided by

automatic processes outside the realm of conscious awareness (Bargh, 1994).1.2 In the last twenty

years or so, experimental social psychologists have demonstrated that the automaticity of action

extends beyond inborn reflexes and overlearned motor programs to complex and meaningful

behavior (for reviews, see Bargh & Chartrand, 1999; Ferguson & Bargh, 2004). A central

conclusion in this research is that people can automatically initiate behavioral responses to the mere

perception of a relevant stimulus. For instance, merely encountering positive or negative stimuli can

automatically elicit affective responses such as mood changes (Chartrand, Van Baaren, & Bargh,

2006) evaluations (Fazio, Sanbonmatsu, Powell, & Kardes, 1986) or approach and avoidance

tendencies (cf. Gray, 2001; Neumann, Förster, & Strack, 2003).

Automatic processes may allow people to function relatively efficiently and without much

conscious effort. However, in some situations, people may need to override automatically activated

responses. For instance, after a doctor’s visit, a person may adopt the goal to quit smoking. In these

or similar cases, people experience a conflict between an automatized (e.g., smoking after dinner)

and an intended response (e.g., eating some fruit instead). To resolve such conflict, people need to

exert cognitive control over their responses (Posner & Snyder, 1975; cf. Norman & Shallice, 1986;

Schneider & Shiffrin, 1977).1.3

Cognitive control refers to a process whereby people can override strong but inappropriate

response tendencies. As such, cognitive control can be considered a non-automatic way of

responding that increases people’s behavioral flexibility when automaticity proves too rigid. In


social psychology, cognitive control processes have thus far received relatively little attention (but

for exceptions, see Bayer, Ferguson, & Gollwitzer, 2003; Mischel, Cantor, & Feldman, 1996).

Moreover, to the extent that social psychologists have studied cognitive control, they have been

mostly interested in the failure of cognitive control (Baumeister, Heatherton, & Tice, 1994) or in

situations in which control is illusory (Wegner, 2002). In other areas of psychology, however,

cognitive control has been extensively examined. For instance, the study of cognitive control is a

key issue in cognitive psychology (Gray, 2001), cognitive neuroscience (Ridderinkhof, Ullsperger,

Crone, & Nieuwenhuis, 2004), developmental psychology (Nigg, 2001), and clinical psychology

(Cohen, Braver, & O’Reilly, 1998).

A classic paradigm that has been used to study cognitive control processes is the so-called

Stroop task (Stroop, 1935; for a review see MacLeod, 1991). In a typical version of the task,

participants are asked to name the ink color in which a target (i.e., a word or a meaningless letter

string) is displayed. Participants are further instructed to ignore any meaning of the target words.

Task performance generally becomes more difficult when the meaning of a word refers to a color

that is incongruent with the ink color (e.g., RED displayed in blue) compared to when word

meaning and ink color are congruent (e.g., RED displayed in red), or when the target is a neutral

letter string (e.g., XXXX displayed in blue). A likely reason is that reading the word meaning is a

highly automatic action tendency that conflicts with the task instructions to concentrate on the ink

color. Failure to control the automatic reading impulse leads to increased error rates and slower

responses on incongruent compared to congruent or neutral targets. The amount of such “Stroop

interference” can be regarded as an indicator of how well participants were able to exert cognitive

control.

Intentional Action Control

Recent research suggests that success at cognitive control depends on two separate factors

(De Jong, Berendsen & Cools, 1999; Kane & Engle, 2003). The first factor pertains to the mere


capability to inhibit an undesired impulse. For the ex-smoker, for instance, the critical question is

whether s/he is able to suppress the urge to lighten a cigarette when s/he feels the craving for

nicotine (cf. Tiffany, 1990). A second but related problem is whether the person is able to fully

bring his or her inhibitory capabilities to bear in a given situation (Duncan, 1995). For the ex-

smoker, the critical question is whether s/he has really committed him- or herself to quit smoking

and whether this decision becomes sufficiently accessible when the craving arises. To fully utilize

their inhibitory capabilities, people thus first need to form and actively maintain an internal

representation of how to respond in a critical situation. In other words, people need to form and

maintain an intention.

Intentions can be seen as arrangements that people make with themselves about what to do

in a particular situation. According to traditional views in social psychology, people form intentions

to enact a particular behavior when they think that it is desirable to perform the behavior and when

they think that it is feasible to do (Ajzen, 1988; Fishbein & Ajzen, 1975). This traditional view is

limited for at least two reasons. First, people may form intentions even when the perceived

feasibility of the anticipated action is low (cf. Shah & Higgins, 1997), or people may not form

appropriate intentions even when the perceived feasibility and desirability of an action are high (cf.

Gollwitzer, 1993). Second, and more important in the present context, traditional views have

remained silent about the processes that translate intentions, once they have been formed, into

action (cf. Eagly & Chaiken, 1993). Understanding these processes is important, however, because

people are not always able to implement their intentions successfully.

The psychological processes that lead to the formation and implementation of intentions (or

to its disengagement) have been subsumed under the term action control (cf. Gollwitzer, 1996,

1999; Kuhl, 1984; Kuhl & Goschke, 1994a). According to theories of action control, intentions are

mental representations of action plans that cannot be enacted immediately but have to be postponed

until a future moment. In contrast with merely wishing to exert some action, intentions imply that a


person has committed him- or herself to actually do it (“I really will do it”). After an intention has

been formed, it may persist in an implicit format outside of conscious awareness until it

automatically becomes reactivated when appropriate situational cues have been encountered (cf.

Gollwitzer, 1993; Goschke & Kuhl, 1996; Martin, Kliegel, & McDonald, 2003). However, when

the intended action is complex, difficult, or must not be forgotten, intentions remain represented in

an explicit state that consumes attention (cf. Anderson, 1983). The psychological system that

maintains explicit representations of intentions activated in memory has been referred to as

intention memory (Kazén & Kuhl, 2005; Kuhl & Kazén, 1999).

Explicit intentions have preferred access to consciousness compared to non-intention-related

information (cf. Zeigarnik, 1927). Evidence for such an “intention superiority effect” comes from a

series of studies, in which participants had to learn short scripts of activities (e.g., setting a table)

that were moderately complex because they included several substeps (e.g., distribute the cutlery)

(Goschke & Kuhl, 1993; cf. Marsh, Hicks, & Bink, 1998). After the learning phase, participants

were informed that they would have to execute one of the scripts later during the experiment. In a

subsequent word recognition task, participants had to indicate whether a word belonged to any of

the scripts. Reflecting the intention superiority effect, participants were found to be faster in

recognizing words from the to-be-executed script compared to words from scripts that did not have

to be executed. These findings support the notion that intention memory plays an important role in

the regulation of action (cf. De Jong et al., 1999; Kane & Engle, 2003).

Working Memory

Intention memory is part of a greater system of memory processes that are commonly

referred to as working memory (Baddeley, 1986; Baddeley & Hitch, 1974; Engle, Kane, &

Tuholski, 1999; cf. Miyake & Shah, 1999; see Kazén & Kuhl, 2005, on the subtle functional

differences between intention memory and working memory). The main task of working memory is

the on-line maintenance and manipulation of explicit information in the service of planning,


complex reasoning, and intentional action. This task includes storage components and attentional

processes. Due to capacity constraints, working memory can only process a limited amount of

information at a time (cf. Neumann, 1987; Pashler, 1994). To use its limited capacity efficiently,

working memory has to be frequently updated such that irrelevant information becomes erased,

relevant information becomes maintained and new information that may be potentially relevant,

gains access to working memory. How exactly working memory updates its contents has not yet

been fully understood. However, there is widespread agreement that updating of working memory

relies on the availability of positive affect and its neurobiological substrate dopamine (cf. Braver &

Cohen, 2000; Cohen, Braver, & Brown, 2002; Dreisbach & Goschke, 2004; Dreisbach et al., 2005;

Kuhl, 2000).

A classic way to measure working memory capacity is the operation span (OSPAN) task

(Turner & Engle, 1989; cf. La Pointe & Engle, 1990). In the OSPAN task, participants are required

to calculate simple arithmetic equations while retaining short series of words in memory for later

recall. Working memory capacity is reflected by a person’s span, i.e. the number of words that can

be recalled correctly while performing the calculations. Using various kinds of span tasks (e.g.,

Daneman & Carpenter, 1980), research has shown that there exist substantial individual differences

in working memory capacity (for a review, see Feldman Barrett, Tugade, & Engle, 2004). Relative

to individuals with small working memory capacity, individuals with large working memory

capacity display greater resistance to distraction (Conway, Cowan, & Bunting, 2001), better

academic achievement (Gathercole & Pickering, 2000), and higher general fluid intelligence

(Engle, Tuholski, Laughlin, & Conway, 1999).

A recent set of studies indicates that working memory capacity contributes to cognitive

control. In these studies, the researchers examined Stroop interference among individuals with high

versus low working memory capacity (Kane & Engle; 2003; cf. Long & Prat, 2002; Miyake,

Friedman, Emerson, Witzki, & Howerter; 2000). The ratio of congruent Stroop trials was


experimentally manipulated such that there were either 75% congruent trials or 0% congruent trials

(cf. Logan & Zbrodoff, 1979). The underlying idea was that a large proportion of congruent trials

would encourage participants to neglect the task intention to concentrate on the ink color because

reading the word meaning would lead to correct responses for the majority of trials. By contrast, the

absence of congruent trials was thought to discourage people from neglecting the task intention. In

line with this assumption, Stroop interference was greater in the 75% condition compared to the 0%

condition. More important, however, Stroop interference in the 75% condition was greater for low

capacity individuals than for high capacity individuals. These findings suggest that sufficient

working memory capacity is required to maintain an explicit intention in memory during cognitive

control.

Working Memory under High Demands

People may not always be able to fully utilize their working memory capacity in a given

situation (cf. Ashcraft & Kirk, 2001; Schmader & Johns, 2003). Indeed, sustained use of working

memory can make people less able to utilize their working memory capacity. For instance,

deliberate attempts at controlling one’s thoughts leads to performance decrements in subsequent

reasoning and comprehension tasks (e.g., Schmeichel, Vohs, & Baumeister, 2003; cf. Baumeister &

Vohs, 2004; Muraven & Baumeister, 2000). Furthermore, sustained cognitive control leads to

performance impairments in the same task (Lorist, Klein, Nieuwenhuis, De Jong, Mulder, &

Meijman, 2000), or in subsequent tasks that rely on working memory (Van der Linden, Frese, &

Meijman, 2003). To the extent that working memory is involved in thought suppression and

cognitive control, these findings indicate that sustained use of working memory can reduce its

effectiveness. In the present dissertation, I refer to conditions of sustained working memory use as

high demanding. By contrast, conditions that are not characterized by sustained working memory

use I refer to as low demanding.1.4


High demands may decrease the effectiveness of working memory in a wide range of

everyday life situations that are characterized by, for instance, the activation of multiple conflicting

goals (Kuhl & Helle, 1986; Shah, Kruglanski, & Friedman, 2003), demanding relationships

(Baldwin, Carrell, & Lopez, 1990), pending assignments (Higgins, 1987), performance pressure

(Beilock, Kulp, Holt, & Carr, 2004), or compelling time constraints (De Dreu, 2003). In all these

situations, working memory remains activated for some time.1.5 In line with this reasoning, high

demands have been found to impair subsequent performance on tasks that require high effectiveness

of working memory but not on tasks that can be performed on an automatic or routine level

(Beilock et al., 2004; Boksem, Lorist, & Meijman, 2005).

A possible reason why high demands can reduce the effectiveness of working memory is

because high demands impair the updating function of working memory. As noted above, the

updating function makes working memory more efficient because it helps to erase irrelevant

information and to maintain relevant information. When the updating function is impaired, working

memory operates inefficiently, which in turn reduces its effectiveness (cf. Eysenck & Calvo, 1992).

Taken together, high demands can impair working memory because they reduce the efficiency of

working memory to update its contents.

Coping with Demands: The Role of Action versus State Orientation

A frequently held position in contemporary social psychology is that people are relatively

passive respondents to external influences rather than active and self-determined agents (however,

see Ryan & Deci, 2004). This view is bolstered by observations that human behavior is greatly

determined by automatic processes and routines (Bargh, 1994; Bargh & Chartrand, 1999), and that

perceptions of control can be illusory (Wegner, 2002). In addition, people’s actual control over their

thoughts and behavior often seems to be suboptimal due to capacity constraints (e.g., Dijksterhuis,

Bos, Nordgren, & Van Baaren, 2006; cf. Pashler, 1994), or may even break down completely due to

working memory’s tendency to lose its effectiveness after sustained use (Schmeichel et al., 2003;


Van der Linden et al., 2003). In light of these limitations, one may indeed wonder whether people’s

active agency is restricted to the scarce moments in life that are relatively low demanding.

A rather different view on human agency emerges from action control theories (Gollwitzer,

1993; Kuhl, 1984). A fundamental assumption of action control theories is that people are often

able to mobilize intentional control processes when they encounter hindrances to perform an action

(cf. Ach, 1910; James, 1890). Because high demands can be an important hindrance during goal

pursuit, they may lead people to shield their intentions more intensively. However, high demands

may not invariably promote intensified shielding. According to personality systems interactions

(PSI) theory (Kuhl, 2000, 2001), an elaboration of Kuhl’s (1984) original action control theory,

reactions under high demands depend on whether people are in a meta-static (change-promoting),

or a cata-static (change-preventing) regulatory mode. When in the meta-static mode, people are

decisive and active to pursue intentional action. A meta-static mode is therefore referred to as

action orientation. When in the cata-static mode, however, people are indecisive and hesitant to

change their mental and behavioral states. The cata-static mode is therefore referred to as state

orientation.


Table 1.1 The Demand-Related Subscale (AOD) of the ACS-90 (Kuhl, 1994b; Action-Oriented Responses are Marked with an Asterisk)

1. When I know I must finish something soon: A. I have to push myself to get started* B. I find it easy to get it done and over with 2. When I don’t have anything in particular to do and I am getting bored: A. I have trouble getting up enough energy to do anything at all B. I quickly find something to do* 3. When I am getting ready to tackle a difficult problem: A. It feels like I am facing a big mountain that I don’t think I can climb B. I look for a way that the problem can be approached in a suitable manner* 4. When I have to solve a difficult problem: A. I usually don’t have a problem getting started on it* B. I have trouble sorting out things in my head so that I can get down to working on the problem 5. When I have to make up my mind about what I am going to do when I get some unexpected free time: A. It takes me a long time to decide what I should do during this free time B. I can usually decide on something to do without having to think it over very much* 6. When I have work to do at home: A. It is often hard for me to get the work done B. I usually get it done right away* 7. When I have a lot of important things to do and they must all be done soon: A. I often don’t know where to begin B. I find it easy to make a plan and stick with it* 8. When there are two things that I really want to do, but I can’t do both of them: A. I quickly begin one thing and forget about the other thing I couldn’t do* B. It’s not easy for me to put the thing that I couldn’t do out of my mind 9. When I have to take care of something important but which is also unpleasant: A. I do it and get it over with* B. It can take a while before I can bring myself to do it 10. When I am facing a project that has to be done: A. I often spend too long thinking about where I should begin B. I don’t have any problem getting started* 11. When I have a boring assignment: A. I usually don’t have any problem getting through it* B. I sometimes just can’t get moving on it 12. When I have an obligation to do something that is boring and uninteresting: A. I do it and get it over with* B. It usually takes a while before I get around doing it


Table 1.2 Correlations between Action versus State Orientation (AOD) and Related Personality Variables (Ns Appear between Parentheses)

Variable Correlations r (N) Reference

AOT (Kuhl, 1994b) .21* (123) Jostmann (2006)

BAS – Drive (Carver & White, 1994) .24* (123) “

BAS – Fun Seeking (Carver & White, 1994) -.10 (123) “

BAS – Reward Responsiveness (Carver & White, 1994) .05 (123) “

BIS (Carver & White, 1994) -.27* (123) “

The Big Five (Goldberg, 1992)

Extraversion .19* (247) Diefendorff (2006)

Agreeableness .20* (247) “

Conscientiousness .46* (247) “

Emotional Stability .38* (247) “

Intellect .25* (247) “

Cognitive Failure Questionnaire (Broadbent et al., 1982) -.45* (95) Kuhl & Goschke (1994b)

Mastery – Achievement Motivation (Elliot & McGregor, 2001) .14 (52) Jostmann & Koole (2006)

Performance – Achievement Motivation (Elliot & McGregor, 2001) -.01 (52) “

RFQ – Prevention (Higgins et al., 2001) .35* (67) Koole (2005)

RFQ – Promotion (Higgins et al., 2001) .42* (67) “

RFQ/Proverb – Prevention (Van Stekelenburg, 2006) .12 (123) Jostmann (2006)

RFQ/Proverb – Promotion (Van Stekelenburg, 2006) .34* (123) “

Self-Esteem (Rosenberg, 1965) .53* (82) Koole & Jostmann (2004)

Working Memory Capacity (Turner & Engle, 1989) -.14 (52)1 Jostmann & Koole (2006)

Note. * p < .05, AOD: Demand-Related Action vs. State Orientation, AOT: Threat-Related Action vs. State Orientation, BIS: Behavioral Inhibition System, BAS: Behavioral Approach System, RFQ: Regulatory Focus Questionnaire. 1 Correlations were obtained under nondemanding conditions.

A person may respond in an action- or state-oriented manner depending on the situational

context. However, there also exist relatively stable individual differences in action versus state

orientation (Kuhl, 1994a, 2000; Kuhl & Beckmann, 1994a). Based on this idea, Kuhl (1981, 1994b)


developed a self-report questionnaire (ACS-90) to assess individual differences in action versus

state orientation.1.6 Each item of the ACS-90 represents a demanding situation and an action- versus

state-oriented way to cope with the situation. The items of the ACS-90 are listed in Table 1.1.

Based on their responses, some individuals can be classified as relatively more action-oriented or

more state-oriented.

Individual differences in action versus state orientation likely originate from early childhood

experiences with demanding situations. Specifically, children who are encouraged by their care-

takers to motivate themselves when they are confronted with obstacles are more likely to develop

an action orientation. By contrast, children whose environments discourage them from motivating

themselves (i.e., overly controlling or neglecting environments), are more likely to develop a state

orientation (cf. Koole, Kuhl, Jostmann, & Finkenauer, 2006; Kuhl, 2000).

Theoretically, action versus state orientation may be an important moderator of how people

respond to high demands. Specifically, high demands may trigger action-oriented individuals to

utilize their psychological resources more efficiently in order to better exert action control. By

contrast, high demands may not have similar effects among state-oriented individuals. Specifically,

state-oriented individuals may fail to utilize their psychological resources more efficiently and,

consequently, may not be able to better exert action control under high demands. The effects of

high demands on action control among action- versus state-oriented individuals are likely reflected

on various dimensions of action control including cognition, affect, and behavioral outcome. Based

on these assumptions, the central hypothesis of the present dissertation is that action- compared to

state-oriented individuals are better able to cope efficiently with the cognitive, affective, and

behavioral consequences of high demands.

Research on Action versus State Orientation

Over the past 25 years, individual differences in action versus state orientation have been

extensively investigated by a number of different researchers in various countries including


Germany, The Netherlands, and the USA. Throughout this research, action versus state orientation

has been found to influence people’s functioning across a broad range of different domains

including health (Bossong, 1998), work (Diefendorff, Hall, Lord, & Strean, 2000), sports

(Heckhausen & Strang, 1988), and education (Diefendorff, Lord, Hepburn, Quickle, Hall, &

Sanders, 1998). Moreover, the effects of action versus state orientation have been found to be

independent from other well-established psychological constructs such as the Big Five personality

dimensions and self-efficacy. An overview of the latter research is provided in Tables 1.2 – 1.3 (for

reviews, see Diefendorff et al., 2000; Koole & Kuhl, in press; Kuhl & Beckmann, 1994a).

Table 1.3 Research Indicating that Action versus State Orientation (AOD) has Effects over and above Alternative Personality Variables

Variable Reference

Achievement Motivation (Elliot & McGregor, 2001) Jostmann & Koole (2006)

AOT (Kuhl, 1994b) Jostmann & Koole (in press, 2006); Koole & Jostmann

(2004)

The Big Five (Goldberg, 1992) Diefendorff et al. (2000)

Goal Orientation (Dweck, 1986) Diefendorff (2004)

Learned Helplessness (Seligman, 1975) Bossong (1998)

Optimism (Scheier & Carver, 1985) “

Reappraisal & Suppression Tendencies (Gross &

John, 2003)

Koole & Jostmann (2004)

Self-Efficacy (Bandura, 1991) Diefendorff (2004)

Self-Esteem (Rosenberg, 1965) Koole & Jostmann (2004)

Working Memory Capacity (Turner & Engle, 1989) Jostmann & Koole (2006)

Note. AOD: Demand-Related Action vs. State Orientation, AOT: Threat-Related Action vs. State Orientation.

To date, research has uncovered some initial support for the general idea that action-oriented

individuals cope more efficiently with high demands than state-oriented individuals. A first line of

research suggests that action- compared to state-oriented individuals are more skilled at cognitive

shielding as reflected by efficient use of cognitive resources under high demands. In particular,

action-oriented individuals report less intrusive thoughts (Kuhl & Fuhrmann, 1998) and fewer


everyday lapses of attention (Kuhl & Goschke, 1994b) than state-oriented individuals. Action-

oriented individuals have also been found to use more efficient and parsimonious decision-making

strategies under time pressure than state-oriented individuals (Stiensmeier-Pelster, John, Stulik, &

Schürmann, 1989).

A second line of research suggests that action-oriented individuals are more skilled at

affective shielding as reflected by the ability to maintain or restore a positive affective state under

high demands. One indicator of such affective shielding is that, under high demands, action-

oriented individuals maintain better access to their emotional preferences (Baumann & Kuhl, 2005)

and a more positive and autonomous self-image (Koole, 2004) compared to state-oriented

individuals. Moreover, action- compared to state-oriented individuals report higher well-being

(Baumann, Kaschel, & Kuhl, 2005) and less depression (Bossong, 1998) during stressful periods of

their personal lives. Among college students, having an action orientation is correlated with

increments in positive affect and energy over the course of a semester (Brunstein, 2001).

Recently, Koole and Jostmann (2004) proposed that affective shielding among action-

oriented individuals takes place on an intuitive level that may not rely on conscious control. In line

with this proposition, action- compared to state-oriented individuals under high demands were

faster to provide positive responses to negative target words in a task that measured intuitive

affective shielding (i.e., an affective Simon task; De Houwer & Eelen, 1998). Moreover, action-

compared to state-oriented individuals under high demands were faster to detect happy faces in

crowds of angry faces in a face-discrimination task (Öhman, Lundqvist, & Esteves, 2001).

Importantly, action-oriented individuals were not faster to give negative responses to positive

words, or to detect angry faces in crowds of happy faces, respectively. Accordingly, high demands

led action-oriented individuals to shield specifically against negative affect rather than simply

increasing overall task speed.


Finally, a third line of research has found that action- compared to state-oriented individuals

are better able at behavioral shielding as reflected by the ability to act upon their intentions under

high demands. For instance, various field studies have found action-oriented individuals are more

successful than state-oriented individuals in maintaining a healthy diet (Palfai, 2002; cf. Fuhrmann

& Kuhl, 1998), avoiding negative behavior related to alcohol abuse (Palfai, McNally, & Roy,

2002), and enacting health- and study-related intentions (Beswick & Mann, 1994; Bossong, 1998;

Kendzierski, 1990). Furthermore, experimental research indicates that action- compared to state-

oriented individuals have fewer difficulties to initiate new courses of action (Kuhl & Beckmann,

1994b; Dibbelt, 1996). The reported deficits among state-oriented individuals may explain why

state-oriented individuals receive more negative supervisor ratings (Diefendorff et al., 2000; cf.

Diefendorff, Richard, & Gosserand, 2006), and display reduced academic success at college

(Bossong, 1998; Diefendorff et al., 1998), and impaired athletic performance when pressure to

perform well is high (Heckhausen & Strang, 1988).

Taken together, previous research supports the idea that action orientation contributes to

cognitive, affective, and behavioral shielding against high demands. However, several important

questions remain. First, previous research has not systematically measured or manipulated the level

of demand. It is therefore unclear whether the reported effects are due to higher coping efficiency

or, alternatively, to generally higher levels of cognitive capacity and positive affect among action-

oriented individuals. From the present theoretical perspective, however, action-oriented individuals

should only function more efficiently than state-oriented individuals under high demands. Indeed,

under low demands, state-oriented individuals are expected to function as well or even better than

action-oriented individuals (cf. Koole, Kuhl, Jostmann, & Vohs, 2005). Second, previous research

has largely relied on field studies that lack the control provided by strictly experimental settings. As

a result, much remains unknown about the precise mechanisms through which action orientation

facilitates coping with high demands. Third and last, previous research has often failed to use


validated measures of cognitive, affective, and behavioral performance. Consequently, the

interpretation of many reported findings remains ambiguous.

The Present Dissertation

The research reported in the present dissertation provides a systematic and methodologically

rigorous examination of the role of action versus state orientation in coping with high demands.

Throughout the present dissertation, my central hypothesis was that action-oriented individuals

cope more effectively with high demands than state-oriented individuals. I investigated this

hypothesis using various operationalizations of demands, and across cognitive, affective, and

behavioral measures. More specifically, I examined the impact of action versus state orientation

varying demands on working memory capacity in Chapter 2. In Chapter 3, I examined the role of

action versus state orientation in regulating aversive affective responses, a skill that is relevant to

coping with demanding situations, which often involve worsened affect (Koole & Jostmann, 2004).

In Chapter 4, I examined the impact of action versus state orientation varying demands on

behavioral efficiency, i.e., on Stroop interference.

Level of demand was experimentally manipulated in various ways including visualization

procedures (Chapter 2; cf. Baldwin & Sinclair, 1996), the visual presentation of angry faces outside

of participants’ conscious awareness (Chapter 3; cf. Chartrand & Bargh, 1996), sustained

performance on difficult tasks (Chapter 4; cf. Lorist et al., 2000), or the activation of intention

memory (Chapter 4; cf. Kuhl & Helle, 1986). I also measured level of demand in participants’

personal lives in one study (Chapter 4; cf. Baumann et al., 2005).

Chapter 2 examined the utilization of working memory capacity as a function of different

levels of demand and action versus state orientation. Working memory is a key memory system that

is involved in cognitive control (Kane & Engle, 2003). In order to exert cognitive control under

high demands it is therefore important that people can utilize their working memory efficiently even

in demanding situations. Working memory was assessed by means of an OSPAN task (Turner &


Engle, 1989), and an intention memory task (Goschke & Kuhl, 1993). The central hypothesis in

Chapter 2 was that action-oriented individuals utilize their working memory more efficiently under

high demands, whereas state-oriented individuals utilize their working memory more efficiently

under low demands as indicated by respectively higher span scores and greater intention memory.

People’s basic affective reactions can trigger corresponding behavioral responses (Gray,

2001; Neumann et al., 2003). Because such responses may be incompatible with people’s

intentions, it is important that people are able to shield against undesired affective reactions.

Previous research has shown that action- compared to state-oriented individuals are better skilled at

affective shielding, especially under high demands (Koole & Jostmann, 2004). Building on this

research, Chapter 3 examined whether action- compared to state-oriented individuals are also more

effective at affective shielding when affective responses are triggered subliminally, i.e., outside of

conscious awareness. Subsequent to a subliminal affect induction (Chartrand et al., 2006),

participants indicated their current affective states on an implicit affect measure (Tesser, Millar, &

Moore, 1988). The hypothesis was that action- compared to state-oriented individuals displayed a

more positive affective state after subliminal priming with negative affect.

The ability to override inappropriate action tendencies is vital for intentional action. It is

therefore important that people are able to exert cognitive control even under high demands.

Chapter 4 examined the role of varying levels of demand and action versus state orientation on

cognitive control using Stroop tasks. The central hypothesis in Chapter 4 was that high demands

lead to greater cognitive control among action- compared to state-oriented individuals. Furthermore,

following Kane and Engle (2003), the number of congruent Stroop trials was manipulated in order

to examine the role of working memory in cognitive control. The hypothesis was that action-

compared to state-oriented individuals display greater cognitive control when the number of

congruent Stroop trials is high, i.e., when cognitive control is especially dependent on working

memory.


The reader may observe some redundancies throughout the chapters in terms of theoretical

background and methodology. These redundancies are deliberate because they allow each chapter

to be read and understood separately without the necessity to skip back and forth between different

chapters. In fact, each empirical chapter is based on a published or submitted scientific article. At

the same time, however, it remains important to integrate the various findings and to consider them

in a broader perspective. An initial step towards such theoretical integration is provided in Chapter

5. Specifically, Chapter 5 suggests that more effective coping with high demands among action-

oriented individuals on cognitive, affective, and behavioral levels is due to their greater efficiency

to update working memory. This idea is discussed in relation to contemporary models of working

memory regulation (Braver & Cohen, 2000; Kuhl, 2000).


Chapter 2

Cognitive Shielding

Action Orientation Moderates the Impact of

Demanding Relationship Primes on Working Memory Capacity2.1

The ability to remember the right things at the right time is often critical for achieving one's

goals. For instance, remembering the departure time of one’s train, the name of the restaurant where

one is supposed to meet a business associate, or that one should buy a birthday present for one’s

partner all help to avoid decidedly unpleasant experiences. Unfortunately, various circumstances

such as unexpected phone calls, heavy workloads, or pressing social commitments often make it

necessary to consider many things simultaneously (Shah, Kruglanski, & Friedman, 2003) leaving

train schedules, restaurant names, and birthday presents candidates for unwitting neglect (Reason,

& Mycielska, 1982). To prevent long delays, lost business, disappointed partners, and other

misfortunes, people need to rely on their working memory capacity (Engle, Kane, & Tuholski,

1999).

Working memory capacity is an immensely useful psychological resource that allows people

to remember important information even when they are temporarily distracted. Past research and

theorizing make inconsistent predictions regarding the effects of situational demands on working

memory capacity. One the one hand, people may utilize their working memory capacity less

efficiently under high demands (e.g., Baumeister, 1984). On the other hand, high demands may lead

people to mobilize greater control resources and thereby utilizing their working memory capacity

more efficiently (e.g., Botvinick, Braver, Barch, Carter, & Cohen, 2001). In the present research,

we suggest a possible way to reconcile these conflicting notions. Based on personality systems

interactions (PSI) theory (Kuhl, 2000, 2001), we propose that the effects of situational demands on

working memory capacity are moderated by action versus state orientation. Action-oriented

individuals utilize their working memory capacity more efficiently in high demanding situations,


whereas state-oriented individuals utilize their working memory capacity less efficiently in high

demanding situations. In the following paragraphs, we begin by reviewing the literatures on

working memory capacity and action versus state orientation. We then present two experiments that

tested our theoretical analysis.

How Working Memory Works

Working memory protects stored information against premature loss due to strong

distracters (Baddeley, 1986; Engle et al., 1999; cf. Miyake & Shah, 1999). This vital task involves

both short-term storage and attentional processes. Stored information in working memory is

unlikely to be consciously represented at all times. Instead, such information persists in a state of

heightened activation that makes it explicitly available to conscious awareness. Due to capacity

constraints, working memory can only process a limited amount of explicit information

simultaneously.

Working memory capacity is especially important when people have to remember what to

do. Oftentimes, people have to postpone their intentions because they cannot enact them

immediately (Kvavilashvili, 1987). To the extent that the intended action is novel, difficult, or can

be easily forgotten, the intention needs to remain represented in an explicit format until the situation

is appropriate for enactment. Given that working memory supports such explicit formats (Baddeley,

1986), working memory is intimately involved in the maintenance of explicit intentions in memory

(Kane & Engle, 2003). The latter function is known as intention memory (Kuhl, 2000). There are

some subtle differences between intention memory and working memory (for a discussion, see

Kazén & Kuhl, 2005; Kuhl, & Kazén, 1999). Nevertheless, intention memory and working memory

both rely on the same limited mental resources and are supported by partly overlapping brain

structures (e.g., left prefrontal cortex; cf. Fuster, 1995). In recognition of the important functional

parallels between the two memory systems, we emphasize the similarities between working

memory and intention memory in the present context.


One of the most popular measures of working memory capacity is the Operation Span

(OSPAN) task (Turner & Engle, 1989). In the OSPAN task, participants are required to calculate

simple arithmetic equations while retaining short series of words in memory for later recall.

Working memory capacity is reflected by a person’s span, i.e. the number of words that can be

recalled correctly while performing the calculations. Using various kinds of span tasks, research has

shown that there exist substantial individual differences in working memory capacity. Relative to

individuals with small memory spans, individuals with large memory spans display greater

resistance to distraction, higher general fluid intelligence, and better academic achievement (for a

review, see Feldman Barrett, Tugade, & Engle, 2004).

The Dual Role of Demand

Depending on situational factors, people may not always be able to utilize their working

memory capacity efficiently (cf. Eysenck & Calvo, 1992). Based on previous research and

theorizing (Braver & Cohen, 2000; Kuhl, 2000; Van der Linden, Frese, & Meijman, 2001), we

suggest that people may utilize their working memory capacity less efficiently under conditions of

sustained working memory load. Sustained working memory load may impair the updating function

of working memory such that situation-irrelevant information cannot be erased and new tasks

cannot be properly processed. In the present context, we refer to conditions of sustained working

memory load as high demanding. By contrast, we term conditions without such load low

demanding.

High demands may leave less working memory capacity available in a variety of everyday

life situations or experimental settings such as during the activation of difficult intentions (Kuhl &

Helle, 1986), or multiple incompatible goals (Shah et al., 2003), in the presence of demanding

relationships (Baldwin, Carrell, & Lopez, 1990), or after prolonged engagement in tasks that require

continuous working memory activation (Lorist, Boksem, & Ridderinkhof, 2005). In line with this

idea, previous research has found that people utilize their working memory capacity less efficiently


as a result of stressful life events (Klein & Boals, 2001), mental fatigue (Van der Linden et al.,

2003), ego depletion (Schmeichel, Vohs, & Baumeister, 2003), or when people try to overcome

strong impulses (Zwaan & Truitt, 1998). Evidence thus suggests that working memory is prone to

“choking under pressure” (Baumeister, 1984).

The notion that people better utilize their working memory capacity under low rather than

high demands is intuitively plausible. However, high demands may not always have a negative

impact on working memory capacity. When people encounter high demanding conditions, this may

signal to people that their current amount of working memory capacity is not sufficient. In response,

people may channel more resources to their working memory, thereby utilizing their working

memory capacity more efficiently under high demands. This idea is consistent with recent notions

of top-down control processes. Contemporary theories of action control (Gollwitzer, 1996; Kuhl,

1984; Moskowitz, Li, & Kirk, 2004) assume that people more intensively shield their intentions

against unwitting loss when they experience or anticipate difficulties to enact their intentions.

Furthermore, neuropsychological research indicates that the detection of conflict in information

processing can lead to compensatory strengthening of cognitive control in Stroop-like interference

tasks (Botvinick et al., 2001). Thus, to the extent that working memory capacity is implicated in

action control and cognitive control, people may better utilize their working memory capacity under

high demands.

Regulation of Working Memory: The Case of Action versus State Orientation

It appears that high demands can either undermine or facilitate working memory. As such,

there likely exist moderating variables that determine the direction of the impact of situational

demands on working memory capacity. One such moderator may be the person’s disposition

towards action versus state orientation (Kuhl, 1984, 1994a, 2000; Kuhl & Beckmann, 1994a).

Action orientation is conceived as a meta-static (change promoting) regulatory mode that is

characterized by decisiveness and initiative. Action orientation thus facilitates intentional action. By


contrast, state orientation is conceived as a cata-static (change preventing) regulatory mode that is

characterized by indecisiveness and hesitation. State orientation thus leads to the perseverance of

current behavioral and mental states. In other words, action versus state orientation reflects the

person’s ability to induce the behavioral and mental changes to terminate the status quo (Kuhl,

1984). According to PSI theory (Kuhl, 1994a, 2000), action versus state orientation varies

chronically between individuals, such that some individuals are more action-oriented, whereas

others are more state-oriented.

Action versus state orientation is distinct from classic motivational constructs (Kuhl, 1984).

More specifically, classic motivational constructs refer to people’s preferences to select a particular

class of motivational tendencies. For instance, people may choose to pursue mastery goals or

performance goals (Elliot & McGregor, 2001), adopt a promotion focus or a prevention focus

(Higgins, 1998), or be driven by implicit or explicit needs (McClelland, Koestner, & Weinberger,

1989). By contrast, action versus state orientation refers to the efficiency with which people can

instigate the behavioral and mental changes necessary to enact a particular motivational tendency.

Consequently, action orientation can increase the enactment of any motivational tendencies that

involve making difficult behavioral and mental changes. Consistent with this reasoning, action-

compared to state-oriented individuals are more efficient on both mastery and performance goals

(Diefendorff, 2004), report having more success in both promotion and prevention goals (Koole,

2005)2.2, and are more successful at reconciling their implicit and explicit needs (Baumann,

Kaschel, & Kuhl, 2005). In short, both theoretical and empirical considerations indicate that the

impact of action versus state orientation occurs over and above classic motivation constructs.


Particularly relevant in the present context, action- versus state-oriented individuals have

been found to respond very differently to situational demands. When confronted with increased

demands, action-oriented individuals tend to mobilize their self-regulatory resources. For instance,

relative to their state-oriented counterparts, action-oriented college students are better able to

organize the multiple demands of college life (Diefendorff, Lord, Hepburn, Quickle, Hall, &

Sanders, 1998). In addition, action-oriented individuals better regulate their affective states under

high demands (Koole & Jostmann, 2004; cf. Jostmann, Koole, Van der Wulp, & Fockenberg, 2005;

Chapter 3). Finally, action-oriented individuals maintain or even increase their task performance

under demanding conditions (Heckhausen & Strang, 1988). By contrast, state-oriented individuals

tend to choke under increasing demands (Heckhausen & Strang, 1988). Notably, state-oriented

individuals do not necessarily display performance deficits under low demanding conditions.

Indeed, state-oriented individuals may outperform action-oriented individuals under supportive

conditions, presumably because state-oriented individuals are more receptive to external

motivational support (Koole, Kuhl, Jostmann, & Vohs, 2005).

Given that action versus state orientation generally moderates how people are coping with

demands, it stands to reason that action versus state orientation may also moderate the impact of

demands on working memory capacity. Specifically, the indecisiveness and hesitation that are

characteristic of state-oriented individuals may translate into a tendency to maintain stored

information in working memory even when quick decisions and initiative are more desirable

(Stiensmeier-Pelster, 1994). State-oriented individuals indeed have been found to display a general

tendency to engage their working memory capacity (cf. Goschke & Kuhl, 1993). Whenever an

additional working memory load is externally induced under high demands (e.g., through stressful

life events), state-oriented individuals are likely to remain under such load thereby decreasing the

working memory capacity that is available for subsequent tasks (Jostmann & Koole, 2006; Chapter


4). Consequently, state-oriented individuals may utilize their working memory capacity less

efficiently under high demands.

By contrast, action-oriented individuals are likely to use their working memory capacity

more efficiently with increasing demands. This mobilization pattern is likely due to the decisiveness

and initiative that characterize action-oriented individuals (Kuhl, 1984), which should help these

individuals to reduce the duration and severity of working memory load. Accordingly, when

working memory load is induced under high demands, action-oriented individuals are likely to

terminate this load as soon as task requirements permit. High demands may thus activate action-

oriented individuals’ tendency to update their working memory, thereby making more working

memory capacity available for subsequent tasks (Jostmann & Koole, 2006; Chapter 4; Stiensmeier-

Pelster, 1994). By regularly updating the contents of their working memory, action-oriented

individuals are likely to better utilize their working memory capacity under high demands.

It is important to note that the preceding functional differences between action- and state-

oriented individuals only apply to high demanding conditions. Low demanding conditions put less

strain on working memory, so that the functional advantage of action-oriented individuals will

diminish with lower levels of demand. In fact, some low demanding contexts may even provide

state-oriented individuals with a functional advantage. Such benefits of state orientation are

especially likely to occur when the situational context is rewarding. The experience or anticipation

of rewards is conducive to the updating of information in working memory (cf. Braver & Cohen,

2000; Dreisbach & Goschke, 2004). State-oriented individuals are more receptive than action-

oriented individuals to externally provided incentives (cf. Koole et al., 2005). A rewarding context

may therefore allow especially state-oriented individuals to free up their working memory capacity.

The Present Research and Hypotheses

The present research sought to provide the first systematic empirical test of the idea that

action versus state orientation moderates the impact of situational demands on working memory


capacity. To induce high or low levels of situational demands, we primed a demanding or accepting

relationship schema (Baldwin & Sinclair, 1996; Koole & Jostmann, 2004). Previous research found

that priming mental representations of relationships with a significant other activates a host of

goals, expectations, and self-evaluations that characterize the interaction with that person (Baldwin

et al., 1990; Fitzsimons & Bargh, 2003; Shah, 2003; cf. Chen, Boucher, & Tapias, 2006). Building

on these findings, we reasoned that the activation of relationship schemas would also have

consequences for working memory capacity. Specifically, priming one’s relationship with a

demanding other likely puts a high strain on working memory thereby exhausting its capacity. By

contrast, priming the relationship with an accepting other likely puts a low strain on working

memory. Indeed, to the extent that accepting others provide a rewarding psychological context in

which one feels accepted unconditionally, priming an accepting relationship may even help

individuals to free up working memory capacity (cf. Koole et al., 2005; Koole, Kuhl, Jostmann, &

Finkenauer, 2006). In sum, visualizing a demanding person may induce a psychological context of

high demand, whereas visualizing an accepting person may induce a psychological context of low

demand.

Following the demand manipulation, we administered a set of validated measures of

working memory capacity. In Study 2.1, we measured working memory capacity by means of an

OSPAN task (Turner & Engle, 1989). In Study 2.2, we measured working memory capacity by

means of an intention memory task (Goschke & Kuhl, 1993). Given that working memory capacity

is functionally linked with intention memory (Kazén & Kuhl, 2005), we expected parallel effects on

the intention memory task and the OSPAN task. Specifically, we predicted that action-oriented

individuals would better utilize their working memory capacity under high demands such that

visualizing a demanding relationship leads to higher span scores (Study 2.1), and better intention

memory (Study 2.2) than visualizing an accepting relationship. By contrast, we predicted that state-

oriented individuals’ working memory would choke under high demands such that visualizing a


demanding relationship leads to lower span scores and worse intention memory than visualizing an

accepting relationship.

Study 2.1

Method

Participants and Design

Seventy-four paid volunteers at the Vrije Universiteit Amsterdam (28 men and 46 women,

average age 20) participated in the experiment. The experimental design was 2 (orientation: action

vs. state) x 2 (visualization: demanding vs. accepting) between participants. Participants were

randomly assigned to the demanding (N = 35) or the accepting (N = 39) condition. The main

dependent variable consisted of the number of correctly remembered words during the OSPAN

task.

Procedure

Upon arrival in the laboratory, participants were led into individual cubicles, each

containing a computer. Experimental instructions were administered via a computer-program.

Participants were first informed that they would participate in several unrelated studies that were

allegedly administered together for efficiency reasons. Participants then moved on with the first

study, which was introduced as a study on personality and contained our assessment of individual

differences in action versus state orientation. Next, participants continued with a visualization task

during which either high or low levels of demand were induced. Subsequently, participants

performed the OSPAN task. After this, participants answered a manipulation check and provided

some biographical information. Finally, participants were debriefed, thanked, and paid by the

experimenter.

Independent variables

Individual Differences. To assess individual differences in action versus state orientation, we

used a Dutch translation of the Action Control Scale (ACS-90; Kuhl, 1994b). In both studies, we


administered the demand-related (AOD) and the threat-related (AOT) subscale of the ACS-90.

According to PSI theory (Kuhl, 2001, p. 243; cf. Baumann et al., 2005; Koole & Jostmann, 2004),

demand and threat represent different aversive states each related to different trigger conditions and

different symptoms. AOD measures whether a person is decisive and active (action-oriented) or

indecisive and inertial (state-oriented) under high demands. By contrast, AOT measures whether a

person becomes challenged (action-oriented) or remains threatened (state-oriented) in situations that

are perceived as danger to one’s well-being or self-image (cf. Blascovich & Mendes, 2000). We

found no effects of AOT in the present investigation, which indicates that our manipulation did not

trigger challenge and threat processes, all Fs < 1. We therefore report only effects of AOD. For the

sake of convenience, we use the more general term “action versus state orientation” throughout this

chapter to refer to AOD unless a more precise distinction is required.

The AOD subscale consists of twelve items, which were intermingled and presented in a

different random order for each participant. Each item describes a demanding situation and an

action-oriented versus state-oriented way of coping with that situation. For each item, participants

are asked to select the response that best described their own reaction in that situation. An example

item is “When I have a lot of important things to do and they must all be done soon: (a) I often

don’t know where to begin, or (b) I find it easy to make a plan and stick with it”. In this example,

option a reflects the state-oriented response and option b reflects the action-oriented response.

Action-oriented responses were coded as 1 and state-oriented responses were coded as 0 and

summed for the entire subscale. Reliability for the AOD scale was sufficient (Kuder-Richardson

(KR) 20 coefficient = .78). Scores could range from 0-12. Participants who gave seven or more

action-oriented responses were assigned to the action-oriented group (N = 42); participants who

gave six or fewer action-oriented responses, were assigned to the state-oriented group (N = 37).2.3

Visualization Manipulation. The visualization procedure was modeled after Baldwin and

Sinclair (1996). During the procedure, which was introduced as a “visualization exercise”,


participants were asked to visualize a particular person from their own life. In the demanding

condition, participants were requested to think of a person who was highly demanding of them.

Participants had to type in the initials of this person, which were used throughout the exercise in

referring to the visualization target. Participants were instructed to vividly imagine being with this

person and to re-experience their thoughts and feelings associated with this person. At various

stages during the visualization, participants typed in the experiences that were evoked by the

visualization. At the end of the exercise, participants were asked to rate their ease of visualization.

In the accepting visualization condition, participants went through the same procedure, but instead

visualized a person who was highly accepting of them.

Dependent Measure

Working Memory Task. To measure working memory capacity we used a computerized

version of the OSPAN task adapted from Schmader and Johns (2003; cf. Turner & Engle, 1989).

The OSPAN task interleaves short series of words to memorize with simple mathematical equations

to be evaluated. In the present investigation, each equation began with the multiplication or division

of two positive integers [e.g., 5 x 7]. The product of this operation then had to be added to, or

subtracted from, another positive integer. For each equation, the answer was included in the

expression. The task of the participants was to evaluate by means of a key-press whether the

equation was correct or incorrect [e.g., Is (5 x 7) – 12 = 23?]. When the equation was correct

participants had to press the A-button on the left side of the keyboard, whereas they had to press the

6-button on the numeric pad on the right side of the keyboard when the equation was incorrect.

After each equation, a word to memorize was presented. After a series of equation word pairs (i.e.,

a set) participants were requested to type in all words they recalled from the preceding set. After

this, a new set of equation word pairs started.

Sets differed in length varying from three to five equation word pairs per set. Six sets of

each length were presented, allowing scores to range from 0 to 72. Presentation order of the sets


was randomized for each participant, such that the number of words to recall was unknown until

recall. The generation of mathematical equations followed the criteria of Schmader and Johns

(2003). Half of the equations were correct, the remaining equations were incorrect. The 72 words

used in the test were monosyllabic Dutch nouns that were randomly assigned to sets. Within sets,

equation word pairs were presented in a different random order for each participant. The assignment

of equations and words to sets was identical for all participants.

Each trial within a set began with the appearance of a fixation asterisk on the screen for 1 s

followed by an equation, which remained visible until participants had pressed a response key.

After this, the screen remained blank for 500 ms, followed by the appearance of a word to be

recalled. After 2 s, the screen went blank again for 1 s, followed by the next trial. Sets were

separated by the announcement “next set”, which remained visible for 3 s. The computer

unobtrusively recorded the words recalled, participants’ responses on the equations, and the time

spent on each equation.

Table 2.1 Operation Spans (OSPAN) as a Function of Orientation and Visualization (Study 2.1; Standard Deviations Appear between Parentheses)

Visualization

Participant

Group Accepting Demanding

Action

66.21 (4.21)a

69.00 (1.65)b

State 68.80 (1.82)c

66.05 (4.37)d

a n = 24. b n = 15. c n = 15. d n = 20. Note. Scores could range from 0 (low OSPAN) – 72 (high OSPAN).

Results

Manipulation Check. At the end of the experimental session, participants were asked to rate

on two items how demanding and how accepting the person was whom they had visualized (1 = not


at all, 9 = very much). These items were scored in the same direction and averaged (Cronbach’s

alpha = .88). The visualized person was rated as more demanding in the demanding condition than

in the accepting condition, F(1, 70) = 74.08, p < .001, η p2 = .51 (M = 5.61 vs. M = 2.68). No effects

of action versus state orientation were found on this index, all Fs < 1.

OSPAN Performance. Span scores were assessed by counting the total number of correctly

recalled words. (Redoing the analyses with the number of correctly recalled words only from sets

where all the words in the set were recalled correctly yielded equivalent results; cf. Schmader &

Johns, 2003.) We subjected average span scores to a 2 (orientation: action vs. state) x 2

(visualization: demanding vs. accepting) ANOVA, which yielded a significant interaction, F(1, 70)

= 11.17, p < .002, η p2 = .14. Relevant means are displayed in Table 2.1. Simple effects analyses

revealed that action-oriented participants displayed marginally higher span scores in the demanding

condition than in the accepting condition, F(1, 70) = 5.89, p < .06, η p2 = .08 (M = 69.00 vs. M =

66.21). By contrast, state-oriented participants had significantly lower span scores in the demanding

condition than in the accepting condition, F(1, 70) = 5.31, p < .03, η p2 = .07 (M = 66.05 vs. M =

68.80). Another way to interpret the data is to note that, in the demanding condition, action-oriented

participants had higher span scores than state-oriented participants, F(1, 70) = 6.11, p < .02, η p2 =

.08 (M = 69.00 vs. M = 66.05). In the accepting condition, however, the pattern was reversed, such

that action-oriented participants had lower span scores than state-oriented participants, F(1, 70) =

5.07, p < .03, η p2 = .07 (M = 66.21 vs. M = 68.80).

Supplementary Analyses. Additional analyses revealed that the results remained identical

when we controlled for the number of correctly evaluated equations (Grand M = 88%). Moreover,

we found no effects of orientation or visualization on the number of correctly evaluated equations;

all Fs < 1.


Discussion

As predicted, the effects of situational demands on working memory capacity were

moderated by action versus state orientation. Among action-oriented participants, priming a

demanding relationship context made them utilize their working memory capacity more efficiently

than priming an accepting relationship context as indicated by participants’ OSPAN scores. Action-

oriented individuals thus utilized their working memory capacity especially under demanding

conditions. By contrast, state-oriented participants displayed higher span scores after priming an

accepting compared to a demanding relationship. State-oriented individuals thus utilized their

working memory capacity especially under low demanding conditions. Stated differently, priming a

demanding relationship led to better utilization of working memory capacity among action-

compared to state-oriented participants, whereas the reverse was true after priming an accepting

relationship.

Study 2.2

Study 2.2 examined the effects of situational demands and action versus state orientation on

intention memory. As discussed before, working memory capacity is required for the preparation of

intentional action. Accordingly, a person’s working memory capacity may be reflected in her or his

ability to maintain explicit representations of intentions (Kane & Engle, 2003).

The intention memory task in Study 2.2 differed in several meaningful ways from the

OSPAN task in Study 2.1. The OSPAN task represents a dual-task situation, in which participants

have to memorize information while they are simultaneously processing unrelated information.

Neither type of information is relevant for the preparation of action. By contrast, the intention

memory task did not involve dual-task conditions, and the information processed in the task was

relevant for action preparation. Consequently, extending our empirical analysis to intention memory

provided additional insights into how people utilize their working memory capacity for the purpose

of future action.


To assess intention memory, we used the postponed intention task (Goschke and Kuhl,

1993). In this task, participants have to learn a pair of short scripts describing activities (e.g., “clear

a desk”) that entail several intermediate steps (e.g., “sharpen the pencils”). After participants have

learned both scripts, they are informed that they have to execute one of the scripts later (prospective

script). The other script (neutral script) does not have to be executed by the participants. During a

subsequent recognition task, participants then have to decide whether words presented on the

computer screen had appeared in one of the two scripts. Using this paradigm, past research has

found evidence for an intention superiority effect, i.e. faster recognition latencies for words from the

prospective script compared to words from the neutral script (Goschke & Kuhl, 1993).

Based on the theoretical link between working memory capacity and intention memory (cf.

Kazén & Kuhl, 2005), we predicted that action versus state orientation would moderate the impact

of demands on intention memory performance in Study 2.2 much like it did on OSPAN

performance in Study 2.1. Specifically, we predicted that action-oriented participants would display

a greater intention superiority effect under high demanding conditions than under low demanding

conditions. By contrast, we predicted that state-oriented participants would display a greater

intention superiority effect under low demanding than under high demanding conditions.

Method


One hundred and twenty-six paid volunteers at the Vrije Universiteit Amsterdam (44 men

and 82 women, average age 21) participated in the experiment. The experimental design was 2

(orientation: action vs. state; between participants) x 2 (visualization: demanding vs. accepting;

between participants). Participants were randomly assigned to the demanding (N = 62) or the

accepting (N = 64) visualization conditions. The main dependent variable consisted of participants’

mean recognition times and number of correctly remembered words from the prospective and the

neutral script. Seven participants (5.5% of the entire sample) failed to recognize any words from the


neutral script. Because all participants recognized at least some words of the prospective script,

these seven participants had probably misunderstood the task instructions that they should respond

to words of both scripts. Because it was not possible to measure intention superiority effects in

recognition times for these participants, they were removed from the dataset. One additional

participant (.8% of the entire sample) was excluded because he incorrectly indicated that the neutral

script had to be executed.

Procedure

The equipment and general procedure were similar to Study 2.1. Participants first answered

a few questionnaires including the AOD subscale of the ACS-90 (Kuhl, 1994b) to assess individual

differences in action orientation (KR 20 = .73). Based on their responses, sixty-nine participants

were assigned to the state-oriented group, whereas fifty-seven participants were assigned to the

action-oriented group. Next, participants continued with a practice trial of the postponed intention

task. The scripts used in the practice postponed intention task were different from those used during

the actual postponed intention task. After the practice trial, participants performed the visualization

task. Subsequently, participants went on with the actual postponed intention task during which

responses where recorded for analysis. Next, participants performed some unrelated tasks followed

by a manipulation check. Finally, participants were debriefed, thanked, and paid for their

participation.

Dependent Measure

Postponed Intention Task. The postponed intention task was adapted from Goschke and

Kuhl (1993) and consisted of a learning phase, a distracter task, the presentation of the execution

instruction, and a recognition test. The postponed intention task was described to the participants as

a study on people’s memory for simple activities. Participants were informed that they would

receive descriptions of two simple activities on the computer screen. Participants expected that they

had to learn these descriptions, which were referred to as “scripts”, and that their memory for both


scripts would be tested in a recognition test. Furthermore, it was stressed that participants would

have to execute one of the two scripts later during the experiment. Which of the two scripts would

be executed was randomly determined by the computer.

The scripts consisted each of a script header (e.g., “setting a table”) and four propositions

describing component activities (e.g., “distribute the plates”). The learning phase started with the

script header of the first script appearing on the screen followed by separate presentations of each of

the four component activities. The header and the components of the script were each displayed for

6 seconds. Subsequently, the entire script was displayed for another 30 seconds. After this, the

second script was presented in an identical way. Subsequently, the entire procedure was repeated

such that both scripts were presented two times to the participants. Next, participants proceeded

with a short distracter task during which they had to count backwards in steps of three from a three-

digit number for 45 seconds. Subsequently, participants received the instruction which of the two

scripts they had to execute. The instruction consisted of the word “execute” followed by one of the

script headers (the prospective script). In a second row, the words “do not execute” were displayed

followed by the second script header (the neutral script). This execution instruction was displayed

for 6 seconds. Assignment of the two scripts was counterbalanced between participants such that

each script served equally often as the prospective script and the neutral script respectively.

Immediately after the presentation of the execution instruction, participants received a

recognition test, during which single words were presented on the screen. Participants were required

to indicate for each word whether they had seen that word in one of the two scripts. If the word was

identified as a word from one of the two scripts, participants were to press the A-key on the left side

of the keyboard. When the word was identified as a new word, participants were to press the 6-key

on the numeric pad on the right side of the keyboard. Participants were further told to respond

quickly and accurately and to guess whenever they were not sure about the correct response.

The recognition test started with four warming-up trials. After this, 44 words were presented


sequentially to the participants. Half of the words were derived from one of the two scripts; the

remaining words were new. Half of the new words were semantically related to respectively one of

the scripts. The remaining new words were semantically unrelated to either script. Each word

remained on the screen until participants had given a response. The screen then remained blank for

1.5 seconds. The computer unobtrusively recorded participants’ responses and response latencies.

After the recognition test, participants had to indicate which script had to be executed. Actual

execution of the prospective script was only required during the practice postponed intention task.

Table 2.2 Average Response Latencies (in ms) for Prospective and Neutral Words and Intention Superiority Effect as a Function of Orientation and Visualization (Study 2.2; Standard Deviations Appear between Parentheses)

Visualization

Accepting Demanding

Word Type Word Type

Participant Group Prospective Neutral ISE Prospective Neutral ISE

Action 1401

(387)

1329

(570)

-71

(487)a

1136

(302)

1257

(406)

121

(236)b

State 1241

(347)

1346

(418)

105

(399)c

1314

(399)

1266

(288)

-47

(348)d

a n = 26. b n = 27. c n = 34. d n = 31. Note. ISE = Intention Superiority Effect (Neutral - Prospective).

Results

Manipulation Checks. At the end of the experimental session, participants were asked to rate

how demanding and how accepting the person was whom they had visualized (1 = not at all, 9 =

very much). These two items were scored such that higher scores indicated higher perceived

demandingness of the visualized person (Cronbach’s alpha = .54). Given the low reliability

coefficient, we conducted a 2 (orientation) x 2 (visualization) multivariate analysis of variance

(MANOVA) on participants’ responses on the two items, which revealed a main effect of


visualization, F(1, 117) = 64.93, p < .001, η p2 = .36. On average, the person who was visualized in

the demanding condition was perceived as more demanding than the person who was visualized in

the accepting condition (M = 5.59 vs. M = 3.40). (We report averaged scores to facilitate

interpretation.) The analysis also yielded a main effect of orientation, F(1, 117) = 6.62, p < .02, η p2

= .05. State-oriented participants rated the visualized person as more demanding than action-

oriented participants (M = 4.74 vs. M = 4.14). The interaction between orientation and visualization

was not significant, F(1, 117) = 1.76, p = .19, η p2 = .02.

Recognition Times. Following Goschke and Kuhl (1993), we first eliminated errors (i.e., the

number of script words that were not recognized, 16.2% of all responses) and responses lower than

300 ms (1.6% of all responses). Analyses on recognition latencies were conducted on log-

transformed data to normalize the distribution (Ratcliff, 1993). To facilitate interpretation, we report

untransformed means in ms.

We proceeded by subtracting average recognition latencies for words from the prospective

script from average recognition latencies for words from the neutral script. The resulting difference

scores index the intention superiority effect, that is, better memory for prospective script words

compared to neutral script words. A 2 (orientation: action vs. state) x 2 (visualization: demanding

vs. accepting) ANOVA confirmed the predicted two-way interaction effect between orientation and

visualization, F(1, 114) = 10.59, p < .002, η p2 = .09. Relevant means are displayed in Table 2.2.

Action-oriented participants displayed a larger intention superiority effect in the demanding

condition than in the accepting condition, F(1, 114) = 6.20, p < .02, η p2 = .06 (M = 121 vs. M = -

71). Simple t-tests revealed a significant intention superiority effect among action-oriented

participants in the demanding conditions, t(26) = 2.86, p < .009, but not in the accepting condition, t

< 1. By contrast, state-oriented participants had a smaller intention superiority effect in the

demanding condition than in the accepting condition, F(1, 114) = 4.40, p < .04, η p2 = .04 (M = -47

vs. M = 105). Simple t-tests revealed a significant intention superiority effect among state-oriented


participants in the accepting condition, t(33) = 2.53, p < .02, but not in the demanding condition, t <

1.

Another way of interpreting the (orientation) x (visualization) interaction is to note that, in

the demanding condition, action-oriented participants showed a larger intention superiority effect

than state-oriented participants, F(1, 114) = 4.05, p < .05, η p2 = .03 (M = 121 vs. M = -47). In the

accepting condition, we found a significant reversal of the pattern. Specifically, action-oriented

participants displayed a lower intention superiority effect than state-oriented participants, F(1, 114)

= 6.72, p < .02, η p2 = .06 (M = -71 vs. M = 105).

Supplementary Analyses on Recognition Times. Additional analyses revealed that the effects

on recognition latencies remained intact when we statistically controlled for error rates.

Furthermore, the examination of intention memory effects during the practice task revealed that

action-oriented (M = 83) and state-oriented participants (M = 84) had similar intention memory

effects under such neutral conditions, F < 1. When we included intention memory during the

practice task as a covariate, the (orientation) x (visualization) interaction on intention memory

during the actual task remained significant.

Recognition Accuracy. Following Goschke and Kuhl (1993), we computed from the hit rates

and false-alarm rates a discriminability index A’ that was proposed by Pollack (1970) as a non-

parametric equivalent to d’, and that can be calculated even when the number of observations is

small. Chance performance is reflected by an A’ value of .5, whereas perfect performance yields an

A’ of 1.0. For each participant, we calculated A’ separately for neutral and prospective words.2.4

Discriminability indices A’ were subjected to a 2 (orientation) x 2 (visualization) x 2 (script type:

prospective vs. neutral) ANOVA with repeated measures on the last factor. This analysis yielded

only a significant main effect for Script Type; F(1, 114) = 11.20, p < .002. Specifically, A’ values

for prospective words were higher (A’ = .81) than A’ values for neutral words (A’ = .78). The

analysis revealed no effects of orientation or visualization, all Fs < 1.


Discussion

The results of Study 2 confirmed that action versus state orientation moderates the impact of

situational demands on intention memory. Action-oriented participants displayed a significant

intention superiority effect after priming a demanding but not after priming an accepting

relationship. Among state-oriented participants, however, this pattern was reversed such that state-

oriented participants displayed a significant intention superiority effect after priming an accepting

but not after priming a demanding relationship. The results can also be interpreted separately by

demand condition. After priming a demanding relationship, action-oriented participants thus had a

greater intention memory effect than state-oriented participants, whereas the reverse was true after

priming an accepting relationship.

Notably, intention superiority effects between action- and state-oriented participants did not

differ during the practice task. Because demand was not yet manipulated before the practice task,

this lack of an effect suggests that action- versus state-oriented individuals do not differ in intention

memory under neutral conditions. Moreover, a covariance analysis revealed that that the effects of

demand on intention memory were independent from individual differences in intention memory

under neutral conditions. Of course, such interpretations should be made with caution, given the

inherent difficulties in drawing conclusions from null findings.

Our manipulation checks found that state-oriented participants perceived the visualized

person in both conditions as somewhat more demanding compared to action-oriented participants.

Even though this effect was not anticipated, it could not explain the observed effects of action

versus state orientation on intention memory. First, effects of action versus state orientation on

intention memory were moderated by situational demands, and no such moderation effect was

obtained on the manipulation check. Second, when we included the manipulation check as a

covariate, the (orientation) x (visualization) interaction on intention memory remained significant.


General Discussion

Past research and theorizing have described inconsistent effects of demands on working

memory capacity. On the one hand, research has found that working memory is prone to "choking

under pressure" (Baumeister, 1984), in that people utilize their working memory capacity less

efficiently in demanding situations. On the other hand, recent notions on top-down control

processes suggest that high demands may serve as a mobilization signal to utilize working memory

capacity more efficiently (e.g., Botvinick et al., 2001). The present research suggests that these

apparent inconsistencies can be resolved by considering the role of action versus state orientation as

a moderator of the effects of situational demands. The choking pattern is more characteristic of

state-oriented individuals, who utilize their working memory capacity less efficiently with

increasing demands. The mobilization pattern, on the other hand, is more characteristic of action-

oriented individuals, who better utilize their working memory capacity with increasing demands.

In two studies, action-oriented participants utilized their working memory capacity more

efficiently after visualizing a demanding relationship compared to an accepting relationship. State-

oriented participants displayed the opposite pattern: they had less working memory capacity

available after visualizing a demanding relationship compared to an accepting relationship. Stated

differently, action-oriented participants better utilized their working memory capacity than state-

oriented participants after visualizing a demanding relationship, whereas the reverse was true after

visualizing an accepting relationship. These effects were robust across two very different measures

of working memory capacity, i.e., OSPAN (Study 2.1) and intention memory (Study 2.2). Taken

together, the present research highlights the importance of action versus state orientation in the use

of working memory capacity under varying levels of situational demand.

Why would working memory choke under high demands among state-oriented individuals?

Presumably, state-oriented individuals’ tendency towards indecisiveness and hesitation (Kuhl,

1984) renders them less capable of updating their working memory. Consequently, state-oriented


individuals will be unable to terminate the sustained working memory load that is induced by high

demanding conditions. This, in turn, leaves less working memory capacity available for subsequent

tasks. Under low demanding conditions, working memory is less loaded than under high demanding

conditions. Consequently, it is less relevant to update working memory under low demanding

conditions. Moreover, low demanding conditions can even help state-oriented individuals to update

their working memory, particularly when conditions are rewarding. Visualizing an accepting

relationship likely provided such rewarding context, thereby making more working memory

capacity available for state-oriented individuals to use in subsequent tasks.

Unlike their state-oriented counterparts, action-oriented participants better utilized their

working memory capacity under high compared to low demanding conditions. A likely explanation

for this pattern is that action-oriented individuals have a tendency towards decisiveness and

initiative (Kuhl, 1984), which renders them more capable of updating the contents of their working

memory. Under high demands, action-oriented individuals are more inclined to update working

memory, thereby making more working memory capacity available for subsequent tasks. By

contrast, low demanding conditions do not trigger compensatory updating of working memory

among action-oriented individuals and thus do not lead to more efficient use of working memory

capacity. Moreover, action-oriented individuals are less receptive to rewarding contexts than state-

oriented individuals (Koole et al., 2005). Consequently, visualizing an accepting person does not

necessarily facilitate working memory among action-oriented individuals.

Action-oriented individuals thus utilize their working memory capacity most efficiently

under demanding conditions, whereas state-oriented individuals utilize their working memory

capacity most efficiently under rewarding or accepting conditions. This overall pattern suggests that

having an action orientation is more compatible with a demanding context, while having a state

orientation is more compatible with a rewarding or accepting context. Despite the observed

advantage of state-oriented individuals under rewarding conditions, however, it is important to note


that rewarding or accepting conditions may be often lacking in common situations such as at work

or during one’s study (cf. Diefendorff et al., 1998). Accordingly, action orientation may prove more

advantageous than state orientation in many achievement contexts.

The present research further highlights the role of relationship schemas for the regulation of

basic cognitive processes. Previous work has shown that priming relationship schemas activates

correspondent goal representations and expectations (e.g., Fitzsimons & Bargh, 2003; cf. Chen et

al., 2006). The present findings add to this literature that relationship schemas even influence the

efficiency of basic memory functions such as working memory capacity. The present research thus

reveals an important connection between interpersonal and intrapersonal processes (cf. Vohs &

Finkel, 2006).

Limitations and Avenues for Future Research

The present research has several limitations and thus leaves important issues open for further

investigation. First, unlike previous research (e.g., Van der Linden et al., 2003), we did not find

main effects of demands on measures of working memory capacity. Previous research did not

include measures of action versus state orientation. It is thus possible that previous research has

unwittingly relied on either predominately state-oriented or predominately action-oriented samples.

It is equally conceivable that previous manipulations have induced much different levels of demand

than the present visualization manipulation. For instance, Van der Linden and colleagues (2003)

had participants engage in a demanding task for two hours. It seems likely that even action-oriented

individuals use their working memory capacity less efficiently under such extreme demands.

Another important issue concerns the regulation of working memory capacity under

threatening conditions. Previous research has found that people utilize their working memory

capacity less efficiently as a result of stereotype threat (Schmader & Johns, 2003) and math anxiety

(Ashcraft & Kirk, 2001). However, theoretical considerations suggest that people may be able to

convert feelings of threat into feelings of challenge under conditions that potentially jeopardize their


self-image or well-being (Blascovich & Mendes, 2000). Based on PSI theory (Kuhl, 2001), we

suggest that action versus state orientation on the threat-related dimension (AOT) may be an

important moderator of working memory capacity under threat. Specifically, threat-related action

orientation may lead to the mobilization of working memory capacity under conditions of

stereotype threat or math anxiety, whereas threat-related state orientation may lead to choking under

pressure. Future research may explore this intriguing possibility.

Concluding Remarks

Working memory is a vital psychological function that allows people to remember important

information even when they are temporarily distracted. Past research and theorizing has made

inconsistent predictions whether working memory capacity is utilized more efficiently or less

efficiently under high situational demands. In the present research, we have suggested that these

conflicting notions can be reconciled by considering the role of action versus state orientation. In

particular, action-oriented individuals used their working memory capacity most efficiently under

high demanding conditions, whereas state-oriented individuals used their working memory capacity

most efficiently under low demanding conditions. The present research thus illuminates how

different individuals dynamically regulate the waxing and waning of working memory.


Chapter 3

Affective Shielding:

Subliminal Affect Regulation among Action- versus State-Oriented Individuals3.1

People sometimes feel good or bad without knowing exactly where these feelings come

from. As experimental research suggests, affective reactions can be triggered by subliminal stimuli

that are presented entirely outside of conscious awareness (Zajonc, 2000). Subliminally triggered

affect has been found to exert a pervasive influence on mood, evaluative judgments, social

information processing, and behavior (Chartrand, Van Baaren, & Bargh, 2006; Stapel, Koomen, &

Ruys, 2002; Winkielman, Berridge, & Wilbarger, 2005). Given that affective reactions sometimes

interfere with people’s goal pursuits (Carver & Scheier, 1990), however, it seems important that

people are able to regulate their affective reactions even if these are triggered subliminally. In

Chapter 3, we examine the role of action vs. state orientation (Kuhl, 1981) in the regulation of

subliminally triggered affect. Specifically, we test the notion that action-oriented individuals are

better able to resist the influence of subliminally triggered affect compared to state-oriented

individuals.

Feelings without a Cause: Evidence from Subliminal Priming Research

In recent years, it has become increasingly evident that affective reactions can be based on

only minimal cognitive processing and often even elude people’s conscious awareness. Several

lines of investigation have demonstrated that such “basic affective reactions” (Winkielman et al.,

2005) can be triggered by subtle changes in the environment (Duckworth, Bargh, Garcia, &

Chaiken, 2002), simple body movements (Cacioppo, Priester, & Berntson, 1993), or physical

postures (Stepper & Strack, 1993). Throughout these different kinds of affect inducements, people

typically remain unaware of the source of their affective reactions. Indeed, it appears that basic

affective reactions are often diffuse and free-floating, and “spill over” into people’s experience of

other stimuli that happen to come along (Zajonc, 2000). Moreover, basic affective reactions can


influence behavioral initiative (Hertel & Hardin, 1990), or direct behavioral tendencies towards

approach and avoidance (Chen & Bargh, 1999; Gray, 2001; Neumann, Förster, & Strack, 2003).

Some of the most compelling demonstrations of basic affective reactions can be found in the

literature on subliminal affective priming (for recent reviews, see Berridge & Winkielman, 2003;

Zajonc, 2000). In subliminal affective priming research, participants are first exposed to positive or

negative affective stimuli for very brief durations that preclude conscious detection. After this

initial exposure, participants’ affective reactions are assessed. A pioneering set of studies on

subliminal affective priming was conducted by Murphy and Zajonc (1993). In these studies,

participants were subliminally primed with human faces expressing either happiness or anger.

Following exposure to these subliminal primes, participants had to express their preferences for

unfamiliar Chinese ideographs. Results showed that participants rated ideographs more favorably

when the prime was a happy rather than an angry face. Notably, no affective priming effects

occurred when participants could consciously perceive the happy and angry face primes.

The foregoing paragraphs suggest that the affective system is functionally different from,

and operates independently of, conscious cognitive processing. In line with this notion,

psychological and neurophysiological research has established that basic affective reactions occur

early on in the information processing sequence, and can be triggered with minimal stimulus input

(Berridge & Winkielman, 2003; Zajonc, 2000). Moreover, subliminal affective priming effects

seem to be resistant to attempts at conscious control (Winkielman, Zajonc, & Schwarz, 1997).

The theoretical relevance of subliminal affective priming research extends beyond the

laboratory. In everyday life, people often do not recognize the causes of their affective reactions.

For instance, people show clear affective reactions in response to the weather, even though they

often fail to identify the weather as a cause of their reactions (Schwarz & Clore, 1983). In principle,

people can realize that stimuli such as the weather have influenced their affective states.

Oftentimes, however, the causes of people's affective reactions remain inaccessible to them (Wilson


& Dunn, 2004). People’s minds would possibly be overstrained if the causes for each separate

affective reaction had to be processed on a conscious level. In real life, therefore, it seems likely

that people experience many affective states without knowing their cause. Subliminal affective

priming research may thus provide a window into the nature of basic affective reactions that are

common in everyday life.

Research further suggests that subliminal affective priming can have important downstream

consequences. In many cases, subliminally triggered basic affect does not outlive a second or so

(Murphy & Zajonc, 1993). However, when people encounter sequences of stimuli that are more

consistently positive or negative, subliminal affective priming may accumulate and thereby become

more durable (Chartrand et al., 2006, Stapel et al., 2002; Winkielman et al., 2005). Furthermore, the

effects of subliminal affective priming extend to cognitive and behavioral processes. For instance,

subliminal affective priming has been found to influence stereotyping (Chartrand et al., 2006), as

well as motivational responses such as consumption behavior (Winkielman et al., 2005). Taken

together, these findings suggest that subliminally triggered affect plays an important role in guiding

cognition and behavior.

Action Orientation and Affect Regulation

It is conceivable that basic affective reactions that are triggered by subliminal affective

priming may sometimes impede goal-directed behavior (cf. Carver & Scheier, 1990). To the extent

that basic affect triggers behavioral tendencies (e.g., approach or avoidance) that are incompatible

with people’s current goals, it is important for people to exert some level of control over this type of

affect. For instance, worsened basic affect may activate behavioral avoidance tendencies (Gray,

2001) that interfere with the person’s intention to take initiative (Hertel & Hardin, 1990). Shielding

one’s intentions thus seems to require control over one's basic affective reactions.

How might people control subliminally triggered affect? Given the unconscious nature of

basic affect, it is likely that its regulation also largely operates at unconscious levels. From this


perspective, the control of subliminally triggered affect may be achieved when automatic affective

responses (e.g. dejection caused by a rainy morning) can be overridden in favour of a more adaptive

affective response (e.g., cheering up to get a hard day’s work done). Similar to the cognitive control

of behavior (Norman & Shallice, 1986), the regulation of basic affect requires the activation of

higher cortical functions (e.g., prefrontal cortex) in order to override inappropriate affective

reactions (Kuhl, 2000).

The regulation of basic affect has been addressed by Personality Systems Interactions (PSI)

theory (Kuhl, 2000). According to PSI theory, people are capable of regulating basic affective

reactions, even when they are unaware of their precise causes. This type of affect regulation is

referred to as intuitive affect regulation (Koole & Jostmann, 2004). PSI theory further suggests that

the functional basis for intuitive affect regulation is provided by extension memory, which is

conceived of as a central executive system linked to the prefrontal cortex that provides integrated

knowledge about the self and the environment. If the person responds to worsened basic affect with

the activation of extension memory, affect becomes modulated so that it is congruent with the

person’s current goal intentions, personal motives, and contextual constraints. Because extension

memory provides rapid processing of vast amounts of complex information, its workings do not

necessarily require the involvement of the conscious mind. In principle, intuitive affect regulation

pertains to both positive and negative affect. However, since goal-directed behavior is most

impeded by negative affect (Simon, 1967), intuitive affect regulation is likely to be stronger when

basic affect is worsened rather than when it is improved.

Aided by intuitive affect regulation processes, people can pursue their goals even in the face

of worsened basic affect. However, people do not invariably engage in intuitive affect regulation.

According to PSI theory (Kuhl, 2000), worsened basic affect triggers either a meta-static (change-

oriented) or a cata-static (change-preventing) mode of regulation. During the meta-static mode of

regulation, individuals are likely to activate central executive functions that promote intuitive affect


regulation and goal-directed action. Therefore, the meta-static mode is referred to as action

orientation. By contrast, during the cata-static mode, central executive functions and intuitive affect

regulation are likely to be inhibited. Instead, individuals are more likely to become preoccupied

with their current states while in a cata-static mode. The cata-static mode is therefore referred to as

state orientation. Taken together, action versus state orientation are mutually exclusive regulatory

modes that become triggered under conditions of worsened basic affect.

Whether a person is more likely to become action-oriented or state-oriented in response to

worsened basic affect depends in part on stable individual differences (Kuhl, 2000). Based on their

prior experiences with situations that cause significant changes in basic affect (i.e., demanding or

threatening situations), some individuals may have learned to react in a predominantly action-

oriented manner under conditions of worsened basic affect. By contrast, other individuals may have

developed a tendency to react in a state-oriented manner when basic affect becomes more negative.

In line with this reasoning, Kuhl (1981, 1994b) developed a self-report scale to assess individual

differences in action versus state orientation. Based on their scores, some people can be

characterized as predominantly state-oriented, whereas others can be characterized as

predominantly action-oriented (for reviews and construct validity, see Diefendorff, Hall, Lord, &

Strean, 2000; Kuhl & Beckmann, 1994a).

Research has gathered ample evidence that individual differences in action versus state

orientation are closely linked to intuitive affect regulation. A first line of research has addressed the

physiological and behavioral consequences of intuitive affect regulation. For instance, Heckhausen

and Strang (1988) observed increased physiological response (i.e., lactate concentration) and

decreased athletic performance among state-oriented basketball players under performance

pressure, whereas action-oriented players displayed no such effect in response to performance

pressure. In a similar vein, Kuhl (1981) found performance decrements on a complex cognitive task

after repeated failure experiences among state-oriented but not among action-oriented participants.


Furthermore, a negative mood impaired complex coherence judgments among state-oriented but not

among action-oriented individuals (Baumann & Kuhl, 2002). Finally, Jostmann and Koole (2006;

Chapter 4) demonstrated that the induction of high demands diminished cognitive control, as

indicated by the performance on the Stroop color evaluation task and related paradigms, among

state-oriented individuals but not among action-oriented individuals.

A second line of research has focused more directly on the affective consequences of action

orientation. Specifically, action-oriented individuals reported less unpleasant feelings in response to

repeated failure experiences (Brunstein & Olbrich, 1985). In a similar vein, action-oriented

individuals showed less depressive symptoms compared to state-oriented individuals, especially

when levels of stress were high (Rholes, Michas, & Shroff, 1989). Furthermore, increases in action

orientation predicts whether phobic patients can overcome their phobic fears (Schulte, Hartung, &

Wilke, 1997).

Koole and Jostmann (2004, Study 1) examined the temporal dynamics of affect regulation in

action- versus state-oriented individuals. In line with prior research (e.g., Brunstein & Olbrich,

1985), action-oriented participants displayed significant down-regulation of tense mood after the

induction of mild stress. More important, however, this down-regulation was not so much apparent

immediately after the induction of stress, but rather when moods were assessed 10 minutes

afterwards. No similar decreases in tension were found when no stress was induced or among state-

oriented participants. This finding supports the notion that action orientation entails efficient down-

regulation rather than a decreased sensibility towards aversive affect.

Finally, research has begun to investigate how action orientation regulates basic affective

reactions without relying on explicit affect measures. Specifically, Koole and Jostmann (2004,

Study 2) examined the effects of action orientation in an affective Simon task. In this task,

participants were required to provide a positive or negative response to a target stimulus on the

basis of a non-affective stimulus feature (e.g., grammatical status). Although the valence of the


target stimulus was irrelevant and should be ignored during the task, participants displayed an

“affective Simon effect” (De Houwer & Eelen, 1998), i.e., they gave faster responses and made less

errors when the valence of the target stimulus was congruent with the valence of the response.

However, when the affective Simon task was preceded by the induction of mild stress, action-

oriented participants were faster and made less errors compared to state-oriented participants on

trials that required them to give positive responses to negative targets. This effect was interpreted as

an up-regulation of worsened basic affect evoked by the prior induction of mild stress. Taken

together, these findings suggest that action orientation can regulate basic affective reactions. It

remains to be seen, however, whether action orientation can also regulate basic affective reactions

that are triggered subliminally.


We designed the present research to examine the link between action orientation and

regulation of subliminally triggered basic affect. To address this issue, we used a subliminal

parafoveal priming paradigm (Chartrand & Bargh, 1996) to prime action-oriented and state-oriented

participants with schematic drawings of angry, neutral, or happy human faces (Öhman, Lundqvist,

& Esteves, 2001). These stimuli are displayed in Figure 3.1.

Angry Happy Neutral

Figure 3.1. Experimental Stimuli used as Subliminal Affective Primes

(Adapted from Öhman et al., 2001).

We assumed that the schematic faces have an intrinsic affective value to people that is

similar for action- and state-oriented individuals. To test this assumption, we conducted a pilot

study (N = 59). In this study, action-oriented and state-oriented participants indicated to what extent


they perceived the happy, neutral, and angry faces as being negative, disapproving, and aggressive

(1 = not at all, 9 = very much). Action-oriented and state-oriented participants’ evaluations of the

happy and angry faces did not differ, all Fs < 1. Furthermore, the two groups evaluated the neutral

face similarly as negative and disapproving, Fs < 1. The only statistically detectable difference lies

in the perceived aggressiveness of the neutral face, F(1, 57) = 4.85, p < .05. State-oriented

participants found the neutral face more aggressive than action-oriented participants, M = 5.00 vs.

M = 3.97, respectively. Overall, we concluded that action-oriented and state-oriented individuals do

not differ in their evaluation of the prime stimuli.

In the parafoveal priming task, participants were subliminally exposed to angry, or neutral,

or happy schematic faces. The exposure time for each face was 30 ms, which has been shown to be

below the level of conscious awareness in this task (Chartrand & Bargh, 1996). After the subliminal

priming task, participants completed a preference judgment task to measure basic affect (Tesser,

Millar, & Moore, 1988).

Our main hypothesis was that action orientation would regulate the effects of subliminal

affective priming on basic affective reactions. More specifically, we expected that subliminal

affective primes would evoke congruent reactions among state-oriented participants, such that

subliminal priming of angry faces would lead to lower basic affect compared to subliminal priming

of happy faces. By contrast, we expected no such priming effects among action-oriented

participants. As regards subliminal priming of neutral faces, we expected that basic affective

reactions in response to those faces would lie in between responses to angry and happy faces.

Study 3.1

Method


Ninety-two paid volunteers at the Free University of Amsterdam (33 men and 59 women,

average age 20) participated in the study. The design was 2 (action orientation: action vs. state;


between participants) x 3 (prime valence: happy vs. neutral vs. angry; between participants).

Participants were randomly assigned to the experimental conditions (happy: N = 31; neutral: N =

30; angry: N = 31). The dependent variable consisted of participants’ mean ratings on the basic

affect measure.

Procedure

Upon arrival in the laboratory, participants were led to individual cubicles each containing

an Apple Macintosh (iMac) computer. Experimental instructions were administered via computer-

program. Participants were first informed that they would complete a series of unrelated tasks,

which were ostensibly administered together for efficiency reasons. Participants first completed a

study on “personality”, which contained a Dutch translation of the Action Control Scale (ACS-90;

Kuhl, 1994b). The ACS-90 served as our measure of individual differences in action versus state

orientation. On completion of the ACS-90, participants moved on with a parafoveal priming task, in

which happy, angry, or neutral faces were presented outside of conscious awareness. Participants

were told that very brief flashes would appear at unpredictable places and times, and that their task

was to decide whether a flash appeared on the left side or the right side of the screen. Immediately

after the parafoveal priming task, participants completed a preference judgment task to measure

basic affect. Next, participants moved on to some unrelated filler tasks, which lasted about fifteen

minutes. Finally, participants continued with a funneled debriefing procedure (Chartrand & Bargh,

1996), after which they were thanked for their efforts and paid.

Independent Variables

Individual Differences. The ACS-90 (Kuhl, 1994b) measures two interrelated but

conceptually independent types of action versus state orientation: first, the capacity to initiate goal-

directed action under high demands (demand-related action orientation, AOD), and second, the

capacity to engage in goal-directed action in response to threatening experiences (threat-related

action orientation, AOT). According to PSI theory (Kuhl, 2000, 2001, p. 243), threat and demand


represent qualitatively different types of aversive affective states, each requiring a specific form of

intuitive affect regulation.

In the present research, we administered two 12-item subscales of the ACS-90, which

respectively measure AOD and AOT. The items of both subscales were intermingled and presented

in a different random order for each participant. Each of the items describes a stressful situation,

and an action-oriented versus state-oriented way of coping with that situation. For each item,

participants were asked to select the response that best described their own reaction to that situation.

An example item of the AOD subscale is “When I know that I must finish something soon: A. I

have to push myself to get started. B. I find it easy to get it done and over with”. An example item

of the AOT subscale is “When I am being told that my work is completely unsatisfactory: A. I feel

paralysed. B. I don’t let it bother me for too long.” In both example items, A represents a state-

oriented response, and B represents an action-oriented response. Action-oriented responses were

coded as “1”, whereas state-oriented responses were coded as “0”. Scores were then summed for

each subscale. Participants who gave 7 or more action-oriented responses on the AOD scale were

classified as action-oriented (N = 45), whereas participants with less then 7 action-oriented

responses were classified as state-oriented on that subscale (N = 47). An identical procedure was

followed for the AOT scale, N = 34 vs. N = 58, respectively.3.2

A priori, we had no strong predictions on whether the subliminal priming procedure would

primarily evoke effects of AOD or AOT. To the extent that the angry faces would be perceived as

threatening (Öhman et al., 2001), we would expect to find effects of AOT rather than AOD.

However, the angry faces that were used in the present study had a closed mouth with a downward

curve, whereas threat is much more strongly associated with an open mouth with bared teeth

(Aronoff, Barclay, & Stevenson, 1988). A closed mouth in a downward curve has been found to

signal strong negativity but only moderate activity (Lundqvist, Esteves, & Öhman, 1999). In view

of these considerations, it seems possible that the angry faces would signal disapproval rather than


threat, which would lead one to expect effects of AOD rather than AOT. In the remainder of this

chapter, we will generally refer to the broader term “action orientation” rather than to its specific

subcategories unless a more precise distinction is required for clarification.

Subliminal Affective Priming. The parafoveal priming task was modelled after Chartrand

and Bargh (1996). The priming task was introduced to participants as a study on reaction speed to

visual stimuli. We instructed participants to react as quickly and as accurately as possible to brief

flashes that appeared at various locations on the computer screen. When a flash appeared on the left

side of the screen, participants were to press a green key on the left side of the keyboard. When a

flash appeared on the right side of the screen, they were to press a red key on the right side of the

keyboard. Participants were told that the flashes would appear at unpredictable times and locations,

and that the best strategy would be to keep their eyes on a fixation point in the center of the screen.

Before participants started with the actual task, they were given five practice trials. During the

practice trials, no faces were primed. During the following 24 experimental trials, participants were

subliminally primed with happy, or angry, or neutral faces. Each face prime was presented for 30

ms, and was immediately followed by a 100 ms mask consisting of dots and lines. No feedback was

given to the participants regarding their performance during the parafoveal priming task (for more

details regarding the parafoveal priming task, see Chartrand & Bargh, 1996).

Basic Affect Measure

Our measure of basic affect was adapted from Tesser et al. (1988). Participants were

informed that the next study was on the aesthetic evaluation of words. They were to evaluate the

pleasantness of some non-existing words on a scale ranging from 1 (unpleasant) to 7 (pleasant).

Only the two anchors of the scale were labeled. Participants were told that the experimenters had

chosen non-existing words in order to avoid disturbing influences of word meaning. Participants

were instructed not to spend too much time on the evaluation and to simply report their first


reactions. The three non-existing words (“pleban”, “lempon”, “tokitorer”; Stapel, 2002) were shown

in random order for each participant.

Awareness Check

At the end of the experimental session, we assessed whether participants had been aware of

the priming procedure. We used a funneled debriefing procedure (Chartrand & Bargh, 1996) in

which participants were asked questions about the study with increasing specificity. First,

participants had to indicate what they thought had been the purpose of the tasks they had performed.

Next, they were asked to indicate whether they thought that the tasks were related to each other, and

whether anything about the study seemed strange or suspicious to them. Finally, participants were

asked to indicate what they had seen during the flashes. No participants guessed the real purpose of

the experiment, neither did they report being suspicious about the priming procedure. The majority

of participants (92.4 % of the entire sample) reported to have seen something unrelated to the prime

stimuli (e.g., a pattern of lines). Seven participants (7.6 % of the entire sample), however, reported

that they had seen a head, a face, or a smile during the flashes. These participants were excluded

from further analysis.

Table 3.1 Basic Affect as a Function of Orientation and Subliminal Affective Priming (Study 3.1; Standard Deviations Appear between Parentheses)

Subliminal Affective Priming

Participant

Group Happy Neutral Angry

Action 3.50

(1.29)a 3.71

(1.10)b 4.04

(.63)c

State 4.24

(.85)d 3.44

(.96)e 3.17

(1.37)f

a n = 20. b n = 14. c n = 18. d n = 18. e n = 13. f n = 10. Note. Ratings were made on a scale from 1 = “unpleasant” to 7 = “pleasant”.


Results

Reliability coefficients for AOD (Kuder-Richardson (KR) 20 coefficient = .70) and AOT

(KR 20 coefficient = .73) were satisfactory. We only report effects on AOD, since all effects on

AOT were non-significant (all Fs < 1). Reliability for our basic affect measure was unacceptably

low (Cronbach’s alpha = .21). Therefore, we decided to conduct both multivariate analyses with the

three single implicit affect items as separate dependent variables, as well as univariate analyses with

the averaged overall basic affect scale as a dependent variable.3.3

Both analytic strategies yielded similar results. Specifically, a 2 (action orientation: action

vs. state; between participants) x 3 (prime valence: angry vs. neutral vs. happy; between

participants) multivariate ANOVA with three basic affect scores as dependent variables yielded a

significant interaction between action orientation and prime valence, F(6, 154) = 2.56, p < .03.

Likewise, a 2 (action orientation: action vs. state; between participants) x 3 (prime valence: angry

vs. neutral vs. happy; between participants) univariate ANOVA with participants’ averaged basic

affect ratings as dependent variable revealed a significant interaction between action orientation and

prime valence, F(2, 79) = 4.46, p < .05, η2 = .10. To facilitate the interpretation of the effect,

however, we report only the simple analyses on the averaged overall basic affect scale. Relevant

means are displayed in Table 3.1.

LSD post hoc analyses revealed that prime valence had a significant effect on basic affect

among state-oriented participants, F(2, 79) = 4.40, p < .05, η2 = .10. More specifically, subliminal

priming with happy faces resulted in higher basic affect compared to subliminal priming with

neutral (p < .05), or angry faces (p < .01), M = 4.24 vs. M = 3.44 vs. M = 3.17, respectively.

However, prime valence had an opposite, albeit nonsignificant, effect on basic affect among action-

oriented participants, F(2, 79) = 1.01, p > .2. Subliminal priming with happy faces resulted in

slightly lower basic affect among action-oriented participants compared to subliminal priming with

neutral or angry faces, M = 3.50 vs. M = 3.71 vs. M = 4.04, respectively.


An alternative way to interpret the results is to note that subliminal priming with angry faces

resulted in lower basic affect among state-oriented participants compared to action-oriented

participants, F(1, 79) = 4.76, p < .05, η2 = .06. By contrast, subliminal priming with neutral faces

did not yield any differences in basic affect between state-oriented and action-oriented participants,

F < 1. Finally, subliminal priming with happy faces resulted in marginally higher basic affect

among state-oriented participants compared to action-oriented participants, F(1, 79) = 3.86, p =

.053, η2 = .05.

Additional analyses revealed that performance during the parafoveal priming task had no

effect on basic affect ratings. More specifically, when we reran the analyses with the average

number of incorrect responses during the parafoveal priming task (3.8 %) and participants’ mean

response times for each trial (334 ms) as covariates, all results remained intact.

Discussion

Past research has shown that subliminal affective priming can have a significant impact on

people’s basic affective reactions (Chartrand et al., 2006; Murphy & Zajonc, 1993; Stapel et al.,

2002; Winkielman et al., 2005). In the present chapter, we explored whether action-oriented

individuals are capable of regulating subliminally triggered basic affect. Our results showed that

subliminal affective priming effects were only found among state-oriented participants. By contrast,

action-oriented participants’ basic affective responses were not influenced by subliminal affective

priming. Indeed, action-oriented participants displayed higher basic affect compared to state-

oriented participants in response to priming with angry faces. Taken together, the present research

provides evidence that action-oriented individuals are able to regulate subliminally triggered basic

affect.

The present findings add to prior research that has demonstrated efficient down-regulation

of basic affect among action-oriented individuals (Koole & Jostmann, 2004). In this prior

demonstration, however, the affect eliciting stimuli were consciously accessible to the participants.


The present study is thus the first to show that action orientation promotes affect regulation even

when the source of the affective reaction remains unrecognized. This finding is notable because

prevalent theorizing has suggested that subliminal affective reactions unfold early during stimulus

perception and relatively unimpeded by top-down processes (Zajonc, 2000). Moreover, prior

research has shown that subliminally elicited basic affect resists conscious interventions

(Winkielman et al., 1997).

How can we reconcile the present findings with existing ideas about the nature of

subliminally elicited basic affect? PSI theory (Kuhl, 2000) suggests that action-oriented individuals

regulate their affective states so that affect facilitates goal-directed behavior. Goal-directed behavior

is served best when individuals do not suppress any affective signal that comes along but remain

susceptible to potential opportunities and dangers (Carver & Scheier, 1990). In line with this

assumption, prior research has shown that intuitive affect regulation among action-oriented

individuals does not interfere with automatic vigilance for negative affect (Koole & Jostmann,

2004, Study 3). Consequently, the present finding does not challenge the idea that basic affective

reactions unfold early during stimulus perception (Zajonc, 2000). However, the present finding does

suggest that people are able to exert control over these basic affective reactions. This control over

basic affect is likely to rely on unconscious processes (Kuhl, 2000) rather than on conscious

attribution (Winkielman et al., 1997).

Subliminal affect regulation seems adaptive for several reasons. First, subliminal affect

regulation allows people to control basic affective reactions without knowing their cause. Second,

subliminal affect regulation presumably does not rely on the limited capacity of the conscious mind.

As a consequence, people are likely to regulate affective reactions that are triggered by more than

one single cause. The present research has shown that action-oriented individuals can regulate

affective reactions that are caused by repeated subliminal exposure to the same (aversive) stimulus.

Although it remains to be examined by future research, we can speculate that action orientation may


not only shield against the affective consequences of a rainy morning (Schwarz & Clore, 1983), but

also against the scowling faces caught up in the same morning’s traffic jam.

Limitations and Future Perspectives

The present research is still preliminary and leaves several questions unanswered. First, in

line with PSI theory (Kuhl, 2000) our suggestion was that action-oriented participants actively

regulate their basic affective reactions. Although this assumption has been confirmed in prior

research (Koole & Jostmann, 2004), it was not directly tested in the present investigation. Future

research on subliminal affect regulation should therefore include more direct measures of affective

regulation (e.g., affective Simon task) as well as pretest measures of basic affect.

Second, the present findings differ from prior research, which has consistently found main

effects of subliminal affective priming rather than an interaction between subliminal affective

priming and action orientation (e.g., Murphy & Zajonc, 1993; Stapel et al., 2002). Prior research

has not controlled for individual differences in action orientation, which renders speculations about

its role during these studies difficult. One possible explanation for the inconsistency between

studies is that prior research has unwittingly relied on observations of predominantly state-oriented

samples. This corresponds with the significant main effect of prime valence among state-oriented

participants in the present research.

An alternative explanation would be that the difference between our reseach and prior work

is due to differences in the situational context in which the priming took place. According to PSI

theory (Kuhl, 2000), action-oriented individuals are unlikely to regulate their affective states when

the situational context endorses relaxation. In line with this assumption, basic affective reactions

between action-oriented and state-oriented participants have been found to be similar under relaxing

and accepting circumstances (Koole & Jostmann, 2004). Possibly, the context in which our research

took place endorsed relaxation to a lesser degree compared to prior research. Further research is


required to gauge the conditions under which subliminal affect regulation among action-oriented

individuals takes place.

Concluding Remarks

Feelings sometimes arise without any obvious cause. Such basic affective reactions pose a

challenge to people’s affect regulation skills because they can be unpredictable and resistant to

attempts at conscious control. Using a subliminal affective priming paradigm, the present research

was able to show that action orientation may allow people to shield themselves against the intrusion

of basic affective reactions. This capacity for subliminal affect regulation seems a remarkable

human achievement, which may promote goal-directed action even when affect creeps in unnoticed.


Chapter 4

Behavioral Shielding

Action Orientation Moderates the Impact of High Demands

in Stroop Interference Tasks4.1

One of the most remarkable achievements of cognitive functioning lies in people’s ability to

override strong but inappropriate action tendencies. Without such cognitive control, people would

be unable to adapt to novel or complex situations, resist temptations, or inhibit their impulses

(Norman & Shallice, 1986). Cognitive control thus provides flexibility where pre-programmed and

habitual behavior control proves too rigid.

People do not always mobilize the same amount of cognitive control in every situation.

Accordingly, the question arises how cognitive control itself is brought about and regulated

(Monsell, 1996). The regulation of cognitive control has been historically assigned to the proverbial

homunculus, or “little man in the head”. However, an auspicious way to exorcise the homunculus

emerges if one considers the capacity of cognitive control processes to make compensatory

adjustments in response to high demands, or conditions of sustained working memory use. The

basic idea is that high demands leave a person with insufficient amounts of cognitive control to

override inappropriate action tendencies (cf. Lorist, Boksem, & Ridderinkhof, 2005). Such a

shortage in cognitive control, however, may lead to compensatory adjustments to strengthen

cognitive control (cf. Gollwitzer, 1993; Kuhl, 1984). Accordingly, the regulation of cognitive

control may be closely coordinated with the demands of the situation.

In the present research, we further elaborate on the role of high demands in the regulation of

cognitive control. Based on Personality Systems Interactions (PSI) theory (Kuhl, 1984, 2000; Kuhl

& Koole, 2004), we suggest that the effects of demand on cognitive control may depend on whether

individuals are able to activate the adjustment processes necessary for strengthening cognitive

control. Individuals who can adjust flexibly to demanding situations, or action-oriented individuals,


are likely to display improved cognitive control under high demands compared to individuals who

are incapable of activating appropriate adjustment processes, or state-oriented individuals (Kuhl,

1994a). In the following paragraphs, we begin by reviewing the existing literature on cognitive

control and action versus state orientation. After this, we present four empirical studies that tested

our theoretical analysis.

Mechanisms of Cognitive Control

Cognitive control (Posner & Snyder, 1975) may be defined as an organization of high-level

executive systems that are responsible for the overriding of strongly activated but situation-

inappropriate action tendencies. Cognitive control thus corresponds to what Miyake and colleagues

(2000, p. 57) denote as “one’s ability to deliberately inhibit dominant, automatic, or prepotent

responses when necessary”. Efficiency at cognitive control is conducive to mental health (Cohen,

Barch, Carter, & Servan-Schreiber, 1999) and healthy life styles (Tiffany, 1990). Moreover,

individual differences in age (Braver et al., 2001), intelligence (Duncan, Emslie, Williams, Johnson,

& Freer, 1996), and working memory capacity (Kane & Engle, 2003) are significantly related to the

functioning of cognitive control. Research has further identified some of the key neural systems that

are related to cognitive control, which include dopamine systems, the prefrontal cortex and the

anterior cingulate cortex (Braver & Cohen, 2000; MacDonald, Cohen, Stenger, & Carter, 2000).

One of the most widely studied paradigms for examining cognitive control is the Stroop

color evaluation task (Stroop, 1935; for a review see MacLeod, 1991). In the typical version of the

task, participants are asked to respond to the ink color in which a word is displayed. Task

performance becomes more difficult when the word meaning refers to a color that is incongruent

with the ink color of the word (e.g., red displayed in blue). Reading the word meaning is a strongly

activated action tendency, even when doing so is task-irrelevant and the word meaning should be

ignored. In the case of incongruent color words, the reading tendency interferes with the task

intention to indicate the ink color as indicated by increased error rates and slower response


latencies. The amount of interference during the Stroop task forms a negative indicator of the

amount of cognitive control that people have exerted.

Stroop interference is multiply determined (Kane & Engle, 2003; cf. De Jong, Berendsen, &

Cools, 1999). One set of factors pertains to the strength of the competing action tendency. When the

strength of the competing action tendency exceeds a critical level during incongruent Stroop trials,

performance decrements become more likely. A different set of factors pertains to the accessibility

of the task intention in working memory. In a Stroop task, the task intention includes the instruction

to ignore the irrelevant word color dimension and focus on the ink color dimension instead.

Increased difficulties to maintain a mental representation of the task instruction in working memory

leads to greater response competition between the word color and the ink color dimensions. This

situation is known as goal neglect (Duncan, 1995; cf. Cohen, Dunbar, & McClelland, 1990; De

Jong et al., 1999; Kane & Engle, 2003).

On a neurological level, goal neglect may reflect decreased dopaminergic activity in relevant

brain areas (cf. Braver & Cohen, 2000; Cohen, Braver, & Brown, 2002). Decreased dopaminergic

activity renders the updating of working memory more difficult (Dreisbach et al., 2005) such that

relevant information (e.g., the task intention) cannot be sufficiently maintained while irrelevant

information cannot be erased from working memory. On a phenomenological level, goal neglect is

experienced as insufficient task attention resulting in a weakened goal drive even if the person is, in

principle, capable of performing the task. Notably, goal neglect may leave the ability intact to

retrieve relevant intentions from memory (Luria, 1966). Goal neglect should thus not be equated

with simply forgetting the task intention. Instead, goal neglect can be regarded as difficulty to

initiate cognitive control.

The role of goal neglect in cognitive control has been investigated by manipulating whether

people need to keep the task intention accessible. For instance, in an early study, enhanced

accessibility of the task intention was provided by explicit on-line instructions and resulted in


improved cognitive control (Luria & Tsvetkova, 1964). More recently, cognitive control has been

found to improve when trials were presented at a relatively fast compared to a slower pace (De Jong

et al., 1999). Likewise, blocked presentation of incongruent Stroop trials reduced interference

compared to blocks with both incongruent and neutral or congruent trials (Kane & Engle, 2003; cf.

Logan & Zbrodoff, 1979; Long & Prat, 2002). In the aforementioned cases, cognitive control was

prompted either through explicit instructions or through frequent or rapidly paced activation during

preceding trials. Frequent or rapidly paced prompts presumably bring the person into a “high-

control state” (Botvinick, Braver, Barch, Carter, & Cohen, 2001) that reduces the attentional burden

to keep the task intention accessible. As such, goal neglect contributes to interference especially

when no external prompting is available.

Drawing on previous work (Braver & Cohen, 2000; Kuhl, 2000; Lorist et al., 2005;

Schmeichel, Vohs, & Baumeister, 2003), we suggest that people are especially vulnerable to goal

neglect under conditions of sustained working memory use. The use of working memory during

mental operations is effortful and draws on limited energetical resources, which become depleted

after sustained use (cf. Muraven & Baumeister, 2000). Importantly, sustained use of working

memory during task performance may also decrease dopaminergic activity and thus render further

use of working memory difficult. Because cognitive control relies on working memory (Kane &

Engle, 2003), sustained working memory use may undermine cognitive control. In line with this

reasoning, recent findings suggest that cognitive control decrements during sustained task

engagement are caused by the failure to maintain adequate levels of dopaminergic activity in

relevant brain areas (Lorist et al., 2005). In the present context, we refer to conditions of sustained

working memory use as high-demanding and conditions without sustained use as low-demanding.

Cognitive Control under Demand: The Case of Action versus State Orientation

High demands may impair cognitive control under various task conditions. For instance,

sustained engagement in tasks that rely on working memory renders cognitive control more difficult


during later phases of the same task (Lorist et al., 2005), or during subsequent tasks (Schmeichel et

al., 2003; Van der Linden, Frese, & Meijman, 2003). Another instance of a high-demanding

condition is the activation of an uncompleted intention in working memory. Because uncompleted

intentions often persist in working memory (cf. Goschke & Kuhl, 1993; Marsh, Hicks, & Bink,

1998), they may cause subsequent cognitive control decrements (Kuhl & Helle, 1986; cf. Kazén &

Kuhl, 2005; Kuhl & Kazén, 1999; Shah, Kruglanski, & Friedman, 2003). Sustained task

engagement and persisting intention activation can also occur outside the laboratory. Indeed, high

demands have been found to undermine performance in academic and athletic settings (e.g.,

Beilock, Kulp, Holt, & Carr, 2004).

Is cognitive control therefore confined to relatively carefree moments in life? Perhaps a

more optimistic conclusion is possible. Both informal observations and theoretical considerations

suggest that cognitive control often proceeds remarkably well even under demanding conditions.

Indeed, people sometimes seem to rise to the occasion and mobilize greater levels of cognitive

control when demands are high. In line with these observations, contemporary theories of

intentional action suggest that people are able to better shield their intentions against unwitting loss

when they perceive or anticipate hindrances (Carver & Scheier, 2005; Gollwitzer, 1993; Kuhl,

1984). Applied to cognitive control tasks, this idea suggests that greater difficulties during cognitive

control may signal that the current level of control needs to be enhanced. Accordingly, greater

difficulties during cognitive control may trigger compensatory adjustments (cf. Miller, Galanter, &

Pribram, 1960). Because high demands render cognitive control more difficult (e.g., Lorist et al.,

2005), it stands to reason that high demands can also trigger compensatory increases in cognitive

control.

Demanding conditions may thus exert a complex influence on cognitive control. On some

occasions, high demands undermine cognitive control, whereas on other occasions, high demands

may lead people to mobilize their cognitive control capacities, resulting in increases in cognitive


control. One factor that influences the dynamic mobilization of control resources may be the

person’s disposition towards action versus state orientation (Kuhl, 1984, 1994a, 2000; Kuhl &

Beckmann, 1994a). Action orientation is conceived as a meta-static (change-promoting) regulatory

mode that is characterized by decisiveness and initiative. Action orientation thus facilitates

intentional action. By contrast, state orientation is conceived as a cata-static (change-preventing)

regulatory mode that is characterized by indecisiveness and hesitation. State orientation thus leads

to the perseveration of current mental and behavioral states. In short, action versus state orientation

indexes whether a person is able to achieve the mental and behavioral changes to terminate the

status quo (Kuhl, 1984; cf. Jostmann, Koole, Van der Wulp, & Fockenberg, 2005; Chapter 3; Koole

& Jostmann, 2004).

People may acquire a stable disposition towards action versus state orientation depending on

their socialization experiences (Kuhl, 1994a, 2000; Koole, Kuhl, Jostmann, & Finkenauer, 2006).

Particularly socialization conditions during childhood seem to be important predictors whether a

person becomes more action- versus state-oriented. Environments that encourage children to

motivate themselves when hindrances occur are likely to foster action orientation. By contrast,

environments that impair children’s ability to motivate themselves (i.e., overly controlling or

neglecting environments) are likely to foster state orientation. Through repeated learning

experiences, some individuals may come to react habitually in an action-oriented manner to high

demands, whereas other individuals come to habitually react to demands in a state-oriented manner.

Kuhl (1981, 1994b) developed a self-report scale that assesses individual differences in action

versus state orientation. Based on individuals' responses on this scale, they can be classified as

either predominantly action- or state-oriented.

Action- compared to state-oriented individuals may be better able to induce the mental and

behavioral changes to exert cognitive control under high demands. As outlined above, a particular

problem with cognitive control under high demands is that high demands can reduce the


effectiveness of working memory (cf. Braver & Cohen, 2000; Kane & Engle, 2003; Lorist et al.,

2005). An effective way to cope with the effects of high demands on cognitive control is therefore

to protect one’s working memory against high demands. Recent research (Jostmann & Koole, in

press; Chapter 2) found that action-oriented individuals make more efficient use of their working

memory capacity under high demands than state-oriented individuals as indicated by their

respective performance in an operation span (OSPAN) task (cf. Turner & Engle, 1989). In view of

the fundamental importance of working memory for cognitive control (Kane & Engle, 2003), these

findings suggest that action-oriented individuals may be better able to exert cognitive control under

high demands than state-oriented individuals.

At first sight, action versus state orientation might seem functionally equivalent to chronic

individual differences in high versus low working memory capacity (Feldman Barrett, Tugade, &

Engle, 2004). Specifically, past research has found individuals with high working memory capacity

to display greater cognitive control in Stroop tasks than individuals with low working memory

capacity (Kane & Engle, 2003; Long & Prat, 2002). However, unlike chronic individual differences

in working memory capacity, action versus state orientation refers to how efficiently individuals

can use their working memory capacity under demanding conditions. Action- compared to state-

oriented individuals update their working memory capacity more efficiently under high-demanding

conditions but not under low-demanding conditions (Jostmann & Koole, in press; Chapter 2).

Accordingly, functional differences between action-oriented and state-oriented individuals in

cognitive control emerge primarily under high-demanding conditions. The dynamic role of the

situational context thus distinguishes action versus state orientation from other well-studied

individual differences such as working memory capacity.

Action versus state orientation is also distinct from individual differences in the motivation

to mobilize great efforts (Kuhl, 1984). Specifically, some individuals may be generally more

motivated to mobilize their efforts during challenging tasks than other individuals (cf. Brehm &


Self, 1989). By contrast, action- versus state-orientated individuals do not systematically differ in

how much effort they are willing to invest. Instead, action- versus state-oriented individuals differ

in how efficiently they can use their processing capacities during challenging tasks (Jostmann &

Koole, in press; Chapter 2; cf. Eysenck & Calvo, 1992). In the context of cognitive control, effort

mobilization often invokes subsidiary costs under high demands due to a speed/accuracy trade-off

(i.e., faster responses at the cost of decreased accuracy or vice versa; see Boksem, 2006; cf.

Hockey, 1997). By contrast, because of action-oriented individuals’ greater processing efficiency

under high demands, greater cognitive control among action- compared to state-oriented individuals

is unlikely to involve a speed/accuracy trade-off.

Research on Action Orientation and Cognitive Control

Though largely indirect, empirical research supports the notion that action versus state

orientation moderates performance of complex tasks that presumably require some amount of

cognitive control. In one study, for instance, action-oriented students reported higher rates of goal

attainment in various life domains compared to their state-oriented fellow students (Diefendorff,

Lord, Hepburn, Quickle, Hall, & Sanders, 1998). Another study found higher correlations between

exercise intentions and behavior among action-oriented compared to among state-oriented

individuals (Kendzierski, 1990). Finally, action-oriented individuals were found to be better able

than state-oriented individuals to control the negative behavioral consequences of alcohol-

consumption (e.g., drunk driving; Palfai, McNally, & Roy, 2002), and to control their eating

behavior (Palfai, 2002).

The aforementioned studies did not measure or manipulate demands. However, additional

work suggests that enhanced cognitive control among action-oriented rather than state-oriented

individuals emerges especially under demanding conditions. First, performance pressure led semi-

professional basketball players to perform better when they were action-rather than state-oriented

(Heckhausen & Strang, 1988). The same investigation revealed reductions in physiological arousal


under performance pressure among action-oriented but not among state-oriented players. Second,

time pressure led action-oriented individuals to engage in a more parsimonious and less time-

consuming decision-making strategies compared to state-oriented individuals (Stiensmeier-Pelster,

1994). Finally, high demands left action-oriented participants more likely than state-oriented

participants to switch from an unattractive to a more attractive activity (Kuhl & Beckmann, 1994b),

and to change directions in a motor movement task (Dibbelt, 1997).

Initial indications suggest that functional differences between action- and state-oriented

individuals are due to a goal neglect mechanism. Recall that the ability to overcome goal neglect

requires a person to initiate cognitive control when the appropriate situational context is

encountered. In line with this, action- compared to state-oriented individuals procrastinate less

(Blunt & Pyschyl, 1998) and have fewer difficulties to initiate actions (Dibbelt, 1997; Kuhl &

Beckmann, 1994b). By contrast, action- compared to state-oriented individuals are not necessarily

better at maintaining a difficult task when task initiation is facilitated by external prompts

(Baumann & Kuhl, 2005; Dibbelt, 1997; Fuhrmann & Kuhl, 1998). Though preliminary, these

findings are consistent with a goal neglect account (cf. Jostmann & Koole, in press; Chapter 2).

The available evidence on action orientation and cognitive control, though suggestive,

suffers from important limitations. First, many of the aforementioned studies did not assess or

manipulate the level of demand. Therefore, in many cases, we can only speculate whether

performance differences between action- and state-oriented individuals were indeed due to high

demands, as our theoretical analysis would suggest. Second, the studies that did consider the role of

demand did not rely on established paradigms in assessing cognitive control. Third and last,

previously used paradigms did not directly examine goal neglect processes in cognitive control (cf.

Kane & Engle, 2003). We designed the present research to overcome these limitations.



In the present research, we examined the role of action versus state orientation in moderating

the impact of high demands on cognitive control. Our general hypothesis was that a high-

demanding context would improve cognitive control among action-oriented compared to state-

oriented participants. In a low-demanding context, by contrast, we did not expect action-oriented

participants to display better cognitive control than state-oriented participants. Indeed, state-

oriented participants sometimes outperform action-oriented participants under low-demanding

conditions (e.g., Jostmann & Koole, in press; Chapter 2; cf. Koole, Kuhl, Jostmann, & Vohs, 2005).

Accordingly, we predicted that state-oriented participants under low-demanding conditions would

display higher or equal levels of cognitive control compared to action-oriented participants.

We conducted four studies that each used different operationalizations of high demands. In

Study 4.1, we induced a demanding context by having participants complete a working memory

task. In Study 4.2, we manipulated high versus low demand as a combination of task duration and

trial difficulty in a Stroop task. In Study 4.3, we examined self-reported levels of demand in

participants’ personal lives. In Study 4.4, we induced high versus low levels of demand by

manipulating the activation of an uncompleted intention. In Study 4.4, we also examined whether

the effects of action versus state orientation were, as hypothesized, mediated by goal neglect (Kane

& Engle, 2003). Across all four studies, we used a Stroop task to measure cognitive control.

In the Stroop task, variations in cognitive control can be manifested either in shifted

response time distributions or in changes in the number of erroneous responses. A slow response

latency on a given trial often reflects the increased difficulty to exert cognitive control. By contrast,

an error often indicates complete lapse of cognitive control on a given trial. One factor that

determines the locus of cognitive control variations is the proportion of congruent relative to

incongruent trials. When the task context includes more congruent than incongruent trials, failure to

exert cognitive control on one of the rare incongruent trials is more likely compared to when the


task context includes equal or even greater numbers of incongruent relative to congruent trials (for

an elaborate discussion on this point, see Kane & Engle, 2003). Studies 4.1 – 4.3 included equal

numbers of congruent and incongruent trials. We therefore expected effects on cognitive control to

emerge in response times rather than errors in Studies 4.1 – 4.3. By contrast, in Study 4.4, we

systematically varied the proportion of congruent and incongruent trials. Depending on whether

conditions included more congruent or more incongruent trials, we expected cognitive control

variations to emerge in either response times or errors.

Study 4.1

In Study 4.1, we sought to obtain initial evidence for our hypothesis that action-oriented

individuals display greater cognitive control in a demanding context than state-oriented individuals.

To induce a demanding context, we administered an OSPAN task, a widely investigated task that

draws on working memory (Turner & Engle, 1989). We reasoned that performing the OSPAN task

would put a strain on working memory and thereby induce a high-demanding context. Because

action-oriented participants are better able to cope with high demands, we expected them to display

greater cognitive control after the OSPAN task than state-oriented participants.

A second purpose of Study 4.1 was to test our assumption that the effects of action versus

state orientation occur over and above individual differences in working memory capacity.

Following past research (Engle, Tuholski, Laughlin, & Conway, 1999; Kane & Engle, 2003), we

assessed individual differences in working memory capacity by computing participants’ scores on

the OSPAN task. Past research has found that high-span individuals display less Stroop interference

than low-span individuals (Kane & Engle, 2003; Long & Prat, 2002). The robustness of this effect

was greater when less incongruent Stroop trials were included, presumably because high numbers

of incongruent trials reduce the attentional burden of working memory. Nevertheless, past research

still found significantly less Stroop interference among high- versus low-span individuals when the

rate of incongruent trials was as high as 80% (Kane & Engle, 2003, Experiment 4; cf. Long & Prat,


2002). Because the rate of incongruent trials in Study 4.1 of the present research was considerably

smaller (33%), we expected high-span individuals to display less Stroop interference than low-span

individuals.

To summarize, the OSPAN task in Study 4.1 served two functions. First, the task was used

to induce a demanding context. Second, the task allowed us to measure participants’ working

memory capacity. Recent work has shown that action orientation is only related to higher span

scores under high-demanding conditions (Jostmann & Koole, in press; Chapter 2). Because

participants in Study 4.1 performed the OSPAN task under low-demanding conditions, we expected

that action versus state orientation would not correlate with OSPAN. Moreover, because of the

conceptual separability between action versus state orientation and working memory capacity (cf.

Jostmann & Koole, in press; Chapter 2; Kuhl, 2000), we expected action- compared to state-

oriented individuals to display less Stroop interference even when we statistically controlled for

individuals’ span scores.

Method

Participants

Fifty-seven paid volunteers at the Vrije Universiteit Amsterdam (16 men and 41 women,

average age 20) participated in the experiment. One participant was excluded from the dataset

because he had responded to the mathematical equations during the OSPAN task in less than a

second per equation (Grand M = 8 s), and had solved less than 50% of the equations correctly

(Grand M = 87%). Three additional participants were excluded because they failed a test for

colorblindness. The main dependent variables consisted of participants’ response latencies and

errors during the Stroop task.

Procedure

Upon arrival in the laboratory, participants were led into individual cubicles, each

containing a computer. The experimenter explained that all instructions would be administered via a


computer program and left. Participants were first informed that they would participate in several

unrelated studies that were allegedly administered together for efficiency reasons. Participants then

began with the first study, which was introduced as a study on personality and contained our

assessment of individual differences in action versus state orientation.4.2 Next, participants moved

on with the OSPAN task, which served as an induction of demand and assessed individual

differences in working memory capacity. Following the OSPAN task, participants completed a

Stroop task, which measured cognitive control. Participants were subsequently tested for

colorblindness by having them to detect numbers among patterns of colored dots. Finally,

participants were debriefed, paid, and thanked by the experimenter.

Independent Variables

Action versus State Orientation. We assessed individual differences in action versus state

orientation by a Dutch translation of the demand-related subscale (AOD) of the Action Control

Scale (ACS-90; Kuhl, 1994b).4.3 The ACS-90 has been validated in over 60 published studies (for

reviews, see Diefendorff, Hall, Lord, & Strean, 2000; Koole & Kuhl, in press; Kuhl & Beckmann,

1994a). These studies have been conducted across a wide range of domains such as physiological

arousal, intention memory, work psychology, therapeutic outcomes, athletic performance, and

medicine intake. Notably, the effects of action versus state orientation are independent of self-

efficacy beliefs (Diefendorff, 2004) and achievement motivation (Heckhausen & Strang, 1988; cf.

Footnote 4.2), and occur over and above the effects of the Big Five personality dimensions

(Diefendorff et al., 2000).

The AOD scale consists of 12 items that were intermingled and presented in a different

random order for each participant. Each item describes a demanding situation and an action-

oriented versus state-oriented way of coping with that situation. For each item, participants are

asked to select the response that best describes their own reaction to that situation. An example item

is “When I have to take care of something important but which is also unpleasant: (a) I do it and get


it over with, or (b) It can take a while before I can bring myself to do it”. Option a reflects the

action-oriented and option b reflects the state-oriented response alternative. Action-oriented

responses were coded as 1, and state-oriented responses were coded as 0 and summed for the entire

scale. Scores could range from 0 – 12. Participants who gave seven or more action-oriented choices,

were assigned to the action-oriented group (n = 36); participants who gave six or fewer action-

oriented responses, were assigned to the state-oriented group (n = 17).4.4 Reliability for the AOD

scale was sufficient, Kuder-Richardson (KR) 20 coefficient = .67.

Operation Span. The OSPAN task was adapted from Schmader and Johns (2003; cf. Turner

& Engle, 1989). In the task, participants were requested to memorize short series of words and

simultaneously evaluate simple mathematical equations. Specifically, participants were to indicate

by a key-press whether an equation was correct or incorrect [e.g., “Is (4 x 8) – 11 = 21?”]. When the

equation was correct, participants had to press the A-button on the left side of the keyboard. When

the equation was incorrect, participants had to press the 6-button on the numeric pad on the right

side of the keyboard. After each equation, a word was presented that had to be memorized. After a

series of equation word pairs (i.e., a set), participants were requested to type in all words they

recalled from the preceding set. After this, a new set of equation word pairs started. All participants

were informed that they would receive feedback about their performance at the end of the task.

Sets differed in length from three to five equation-word pairs per set. Five sets of each

length were presented, allowing scores to range from 0 to 60. Presentation order of the sets was

randomized for each participant, such that the number of words to recall was unknown until recall.

The generation of mathematical equations followed the criteria of Schmader and Johns (2003). Half

of the equations were correct whereas the remaining equations were incorrect. The 60 words used in

the test were monosyllabic Dutch nouns that were randomly assigned to sets. Within sets, equation

word pairs were presented in a different random order for each participant. The assignment of

equations and words to sets was identical for all participants.


Each trial within a set began with the appearance of a fixation asterisk on the screen for 1 s

followed by an equation, which remained visible until participants had pressed a response key.

After this, the screen remained blank for 500 ms, followed by the appearance of a word to be

recalled. After 2 s, the screen went blank again for 1 s, followed by the next trial. Sets were

separated by the announcement “next set”, which remained visible for 3 s. The recall phase at the

end of each set lasted until participants pressed a response key. Thus, the presentation of the

equations and the recall phase were participant-paced, whereas the presentation of the words to

remember was computer-paced. The computer unobtrusively recorded the words recalled,

participants’ responses on the equations, the time spent on each equation, and the time spent during

each recall phase.

To assess individual differences in working memory capacity, we added the number of

correctly remembered words for each participant. Subsequently, we divided the sample at the

midpoint (i.e., 56 words). Accordingly, participants who had remembered 56 or less words correctly

(n = 28) were classified as individuals with low working memory capacity, whereas participants

with more than 56 correctly remembered words (n = 25) were classified as individuals with high

working memory capacity. Given the small sample size in the present research we decided not to

exclude participants from the two mid-quartiles of our sample or participants who had incorrectly

evaluated less than 85% of the equations, as past research has done (Kane & Engle, 2003).

Rerunning the analyses without participants who had evaluated less than 85% of the equations

correctly (n = 15) did not substantially change the results.

Stroop Task

During the Stroop task, participants were asked to respond to colored letter-strings. If a

letter-string on the screen was presented in blue ink, participants were to press the 6-key on the

numeric pad of the keyboard. If the word was presented in red ink, participants were to press the A-

key on the keyboard. Some letter-strings were color words (RED or BLUE), whereas others


consisted of a series of Xs. Stimuli were either displayed in red or blue. Participants were instructed

to ignore the meaning of the words and to focus on the ink colors only. In addition, they were

encouraged to respond quickly while being accurate. To promote quick responding, participants

were asked to place their fingers on the response keys throughout the task. We further encouraged

participants to focus on the presentation location of all stimuli on the center of the screen during the

entire task. Before the actual task, participants completed 10 practice trials. After each practice trial,

the computer provided feedback regarding the accuracy of participants’ responses. Subsequently,

the participants proceeded with the actual task, which consisted of 60 experimental trials. No

feedback was provided during the actual task.

Trials were presented successively and in a different random order for each participant. The

actual task included 20 congruent trials (i.e., RED in red, or BLUE in blue), 20 neutral trials (i.e.,

XXXX in red or blue), and 20 incongruent trials (i.e., RED in blue, or BLUE in red). Each trial

started with the presentation of a fixation asterisk for 1 s, immediately followed by the presentation

of a colored letter string. After participants had given a response, the screen remained blank for 2 s,

until the fixation asterisk reappeared to indicate the onset of the next trial. For each trial, the

computer unobtrusively recorded participants’ response latency and response key. The fixation

asterisk and the target color words were presented in the center of a white computer screen (font

type Arial, size 36, resolution 1024 x 768, on 17˚ CRT Iiyama monitors). During the entire task, the

word “red” in black ink and the response key “A” were displayed in the lower left part of the

screen, and the word “blue” in black ink and the response key “6” were displayed in the lower right

part.


Table 4.1a Response Latencies (in ms) and Errors in Color Evaluation as a Function of Orientation and Trial Type (Study 4.1; Standard Deviations Appear between Parentheses)

Trial Type

Congruent Neutral Incongruent SI

Participant Group Latency Errors Latency Errors Latency Errors Latency Errors

Actiona 521

(109)

.39

(.65)

516

(103)

.42

(.60)

584

(155)

1.00

(1.10)

68

(83)

.58

(1.11)

Stateb 506

(66)

.18

(.53)

503

(90)

.29

(.59)

624

(147)

.65

(.93)

121

(86)

.35

(.86)

a n = 36. b n = 17. Note. SI = Stroop Interference (Incongruent – Neutral).

Results

Color Evaluation Latencies. Before analyzing the response latency data, we removed

erroneous responses (3.5% of all responses) and responses faster than 300 ms (1.5% of all

responses) from the dataset. All analyses in Studies 4.1 – 4.4 were performed on log-transformed

latencies. To facilitate interpretation of the results, we report untransformed means in ms

throughout the article.

On average, participants’ responses on neutral trials were faster than responses on

incongruent trials, F(1, 52) = 57.20, p < .001, ηp2 = .52 (M = 512 vs. M = 597). An additional

analysis revealed that average responses on congruent trials were as fast as responses on neutral

trials, F < 1 (M = 516 vs. M = 512). Thus, color evaluation latencies in the Stroop task revealed

evidence for interference, but not facilitation. Throughout the present investigation, some of the

studies revealed Stroop facilitation whereas others did not. This inconsistent pattern may be due to

the use of Xs as neutral stimuli (cf. MacLeod, 1991). To the extent that we did find Stroop


facilitation in the present investigation, it was not moderated by action versus state orientation, all

Fs < 1. We therefore do not report analyses of Stroop facilitation in the present research.

We proceeded by subtracting average response latencies for neutral trials from average

response latencies for incongruent trials. The resulting difference score can be interpreted as an

index of Stroop interference with higher scores indicating more Stroop interference (cf. MacLeod,

1991). (Subtracting response latencies on congruent trials from response latencies on incongruent

trials as an alternative index for Stroop interference revealed equivalent results across Studies 4.1 –

4.4.) A one-way ANOVA on Stroop interference revealed the predicted effect of orientation, F(1,

51) = 4.49, p < .04, ηp2 = .08. Relevant means are displayed in Table 4.1a. As expected, action-

oriented participants displayed smaller Stroop interference than state-oriented participants (M = 68

vs. M = 121).

Errors. We subtracted the number of errors on neutral trials from errors on incongruent trials

to obtain an index of Stroop interference in errors. A one-way ANOVA on Stroop interference in

errors revealed no effect of orientation, all Fs < 1. Further analyses revealed that the effect of

orientation on Stroop interference in response latencies remained intact when we statistically

controlled for Stroop interference in errors. Thus, the effect of orientation was not due to a

speed/accuracy trade-off.

Supplementary Analyses. We further tested the effect of working memory capacity on

Stroop interference. As expected, high-span participants displayed less Stroop interference in

response latencies than low-span participants (M = 59 vs. M = 108); F(1, 50) = 4.56, p < .04, ηp2 =

.08. Relevant means are displayed in Table 1b. Notably, this analysis included participants’

responses on the AOD scale as a covariate, which did not remove the effect of working memory

capacity on Stroop interference. Rerunning this analysis without two participants whose span scores

were more than 2.5 SD below the group mean (Grand M = 55.72; SD = 3.61) left the results intact,

indicating that OSPAN effects on Stroop interference were not due to the influence of outliers.


Action orientation and OSPAN performance thus both predicted Stroop interference. To

examine whether the effect of action versus state orientation on Stroop interference in response

latencies was independent from chronic individual differences in working memory capacity, we

reran the analysis with participants’ span scores as a covariate. As expected, including span scores

as a covariate left the effect of orientation intact, F(1, 50) = 5.50, p < .03, ηp2 = .10. Moreover,

participants’ AOD scores and their span scores (both entered as continuous variables) were

uncorrelated, r = -.14, ns.

Additional analyses revealed that participants on average studied the arithmetic equations

during the OSPAN task for 8 s before they pressed a response key. Moreover, participants needed

on average 11.7 s to type in the words during the recall phase of each set. Equation presentation

time was uncorrelated with span scores (r = -.10, ns), whereas recall time was negatively correlated

with span scores (r = -.36, p < .02). The effects on Stroop interference in response latencies for

action- versus state-oriented individuals and high- versus low-span individuals, remained intact

when we controlled for equation presentation time, recall time, and the number of correctly solved

equations (Grand M = 87%), respectively.

Table 4.1b Response Latencies (in ms) and Errors in Color Evaluation as a Function of Working Memory Capacity (WMC) and Trial Type (Study 4.1; Standard Deviations Appear between Parentheses)

Trial Type


Participant

Group Latency Errors Latency Errors Latency Errors Latency Errors

Low WMCa 539

(114)

.54

(.74)

528

(113)

.46

(.64)

636

(179)

.93

(1.22)

108

(103)

.46

(1.17)

High WMCb 491

(66)

.08

(.28)

494

(76)

.28

(.54)

553

(103)

.84

(.85)

59

(55)

.56

(.87)



Discussion

In Study 4.1, we found initial evidence for our hypothesis that action versus state orientation

moderates cognitive control under high demands. After performing a demanding task (i.e.,

OSPAN), action-oriented participants displayed less Stroop interference in response latencies than

state-oriented participants. Importantly, reduced Stroop interference in response latencies did not

come at the expense of greater Stroop interference in errors. The effect of action orientation was

thus not due to a speed/accuracy trade-off. This pattern of findings is theoretically meaningful,

because speed/accuracy tradeoffs occur when individuals simply increase their efforts to enhance

task performance under high demands (Boksem, 2006). The results of Study 4.1 thus indicate that

demand-induced increases in cognitive control among action-oriented individuals results from

increased efficiency rather than increased efforts.

Study 4.1 further replicated past research (Kane & Engle, 2003; Long & Prat, 2002) by

demonstrating that participants with high working memory capacity displayed greater cognitive

control than participants with low working memory capacity. Importantly, the effect of action

versus state orientation on cognitive control remained intact when we controlled for individual

differences in working memory capacity. Indeed, Study 4.1 yielded no correlation between action

versus state orientation and individual differences in working memory capacity. In sum, Study 4.1

revealed that action versus state orientation contributes to cognitive control above and beyond

individual differences in working memory capacity (cf. Jostmann & Koole, in press; Chapter 2).

Compared to previous research (Kane & Engle, 2003), participants in Study 4.1 generally

obtained relatively high OSPAN scores. Possibly, our OSPAN task was somewhat easier to perform

than comparable tasks during previous research (but see Schmader & Johns, 2003). Because our

version of the OSPAN task was in part participant-paced, one might wonder whether participants

had used idiosyncratic strategies (e.g., extending the equation presentation time) to reduce the

working memory load of the task (cf. Friedman & Miyake, 2004). However, we think that at least


two arguments speak against undue influence of idiosyncratic strategies in the present investigation.

First, as in previous research (Kane & Engle, 2003; Long & Prat, 2002), span scores in Study 4.1

predicted cognitive control in the Stroop task. Close inspection of the data revealed that this relation

did not depend on outliers with extremely low span scores. Second, correlational analyses revealed

that increased processing time during equation presentation and during word recall was unrelated to

greater span scores. Indeed, controlling for equation presentation time and word recall time had no

effects on the relation between OSPAN and Stroop interference. In brief, there is no indication for

undue influence of idiosyncratic strategies in our OSPAN task.

Study 4.2

In Study 4.2, we examined the effects of action versus state orientation under high- versus

low-demanding conditions. To induce varying levels of demand, we changed some parameters of

the Stroop task itself. Performing the Stroop task strains working memory (Kane & Engle, 2003),

which, after prolonged task duration, depletes processing resources and thus renders the task more

demanding (cf. Lorist et al., 2005). Accordingly, performance on trials that occur later during the

Stroop task should be more demanding than performance on trials that occur early during the task.

Based on this assumption, we predicted that action versus state orientation moderates cognitive

control in the Stroop task after prolonged task duration such that action-oriented participants display

greater cognitive control on later trials than state-oriented participants. We predicted no greater

cognitive control among action- compared to state-oriented participants after short task duration

because demands are relatively low during early trials.

Prolonged engagement in the Stroop task not only strains working memory but also

increases task practice. Task practice, in turn, reduces Stroop interference (cf. Roe, Wilsoncroft, &

Griffiths, 1980). To control for potential practice effects, we manipulated the difficulty of Stroop

trials. Half of the trials were regular Stroop trials consisting of colored strings such as those used in

Study 4.1 (hereafter, “regular trials”). The remaining trials (hereafter, “difficult trials”) included a


second task in addition to the Stroop task. This second task required participants to evaluate two

consecutively presented letters. Participants thus had to divide their attention over the Stroop task

and the anticipation of the letter task during difficult trials. We reasoned that difficult trials would

impose an extra load on working memory as they require dual task coordination including the

monitoring and sequencing of behavioral subgoals (cf. Kazén & Kuhl, 2005; Lavie, Hirst, de

Fockert, & Viding, 2004). Because complex tasks should require more practice than simple tasks,

we reasoned that practice effects would occur on regular trials but not (or at least significantly less)

on difficult trials.

In line with our general hypothesis, we predicted that action versus state orientation would

moderate cognitive control after prolonged task duration on difficult trials but not on regular trials.

Specifically, action-oriented individuals were expected to display less Stroop interference than

state-oriented individuals during difficult trials that occur later during the task.

Method


Participants were 40 paid volunteers at the Vrije Universiteit Amsterdam (16 men and 24

women, average age 21). The experimental design was 2 (orientation: action vs. state; between

participants) x 2 (trial difficulty: difficult vs. regular, within participants) x 3 (task duration: low vs.

intermediate vs. high; within participants). Two participants failed the colorblindness test and were

therefore removed from the dataset. The main dependent variables consisted of participants’

average response latencies and errors in a Stroop task.

Procedure

Participants started by filling out some questionnaires, which contained our assessment of

individual differences in action versus state orientation (KR 20 = .64). Based on their responses,

sixteen participants were classified as state-oriented, whereas twenty-two participants were

classified as action-oriented. We subsequently administered the Stroop task, which consisted of


three consecutive blocks of trials. The Stroop task served as our measure of cognitive control, and

also included our manipulations of trial difficulty and task duration. After the Stroop task, we

administered a test for colorblindness. Finally, participants were paid, debriefed, and thanked for

their participation.

Stroop task

Instructions for the Stroop task were identical to those in Study 4.1 with the following

exceptions: Participants were informed that some ink color evaluations, which were referred to as

“Task 1”, would be followed by a second task (“Task 2”). During Task 2, participants had to decide

whether a consecutively presented pair of letters was of the same type or of a different type. When

the letters were of the same type (i.e., either both letters are vowels, or both letters are consonants),

participants were to press the A-key on the keyboard. When the letters were of a different type (i.e.,

one letter is a vowel, and the other letter is a consonant), participants were to press the 6-key of the

numeric pad of the keyboard. At the beginning of each trial, participants were informed whether the

trial consisted only of Task 1, or of Task 1 and Task 2.

The onset of each trial was indicated by the word “next”, which remained visible for 1 s. It

was immediately followed for 1 s by a prompt that informed participants whether the trial consisted

only of “Task 1”, or “Task 1 + Task 2”. Next, participants saw a fixation asterisk for 1 s, followed

by a blank screen for 1 s. Then, a colored letter string appeared and remained on the screen until

participants pressed a response key. When the trial included only Task 1, the screen subsequently

remained blank for 1 s until the word “next” reappeared on screen indicating the onset of the next

trial. However, when the trial also included Task 2, participants’ responses to the colored letter

strings were followed by a blank screen for 1 s. Next, a letter appeared for 50 ms on the screen,

after which it was replaced by a second letter. The second letter remained on the screen until

participants had given a response. Subsequently, the screen remained blank for 1 s, followed by the

word “next” to indicate the onset of the next trial.


Before starting with the actual task, participants completed a block of 10 practice trials

during which feedback was provided on response correctness. No feedback was given during the

actual task. The actual task consisted of 180 trials divided over three blocks of 60 trials. Task

duration was operationalized as the position of the three blocks. Accordingly, task duration was

lowest during the first block, intermediate during the second block, and highest during the third

block. Between every two blocks, a short break (30 s) was included. In each block, all color words

(congruent, neutral, incongruent) and ink colors (blue and red) were equally represented. During

half of the trials of each block, Task 1 was followed by Task 2. During Task 2, half of the letter

pairs were of the same type, whereas the remaining pairs were of a different type. The order in

which the three blocks were presented was identical for each participant. The presentation of the 60

trials within each block varied randomly between participants.

Results

Stroop Effect

Color Evaluation Latencies. Before analyzing response latencies, we removed erroneous

responses (4% of all responses) and responses faster than 300 ms (1.2% of all responses) from the

dataset. On average, responses on neutral trials were faster than responses on incongruent trials,

F(1, 37) = 60.88, p < .001, ηp2 = .62 (M = 642 vs. M = 733). Average responses on congruent trials

were as fast as responses on neutral trials, F < 1 (M = 639 vs. M = 642). Thus, there was

interference but no facilitation. To examine whether our manipulation of trial difficulty was

successful, we conducted a 2 (trial difficulty) x 3 (task duration) x 3 (trial type) repeated measures

analysis on Stroop interference. As indicated by a significant main effect of trial difficulty,

participants’ responses on regular trials were in general faster than participants’ responses on

difficult trials (M = 661 vs. M = 680), suggesting that trials involving secondary task preparation

were generally more difficult than trials that did not require secondary task preparation, F(1, 37) =

20.86, p < .001, ηp2 = .36.


Table 4.2 Response Latencies (in ms) and Errors in Color Evaluation as a Function of Orientation, Task Duration, Trial Difficulty, and Trial Type (Study 4.2; Standard Deviations Appear between Parentheses)

Trial Type


Participant

Group Latency Errors Latency Errors Latency Errors Latency Errors

Short Task Duration (Block 1)

Low Trial Difficulty

Actiona 659

(142)

.18

(.39)

617

(116)

.27

(.55)

712

(171)

.55

(.80)

94

(113)

.27

(1.08)

Stateb 620

(128)

.63

(.41)

680

(128)

.44

(1.26)

882

(271)

.69

(1.20)

202

(174)

.25

(.77)

High Trial Difficulty

Actiona 651

(127)

.05

(.21)

647

(103)

.36

(.73)

758

(168)

.23

(.43)

110

(108)

-.14

(.84)

Stateb 682

(131)

.44

(1.03)

689

(116)

.56

(1.26)

790

(197)

.63

(.96)

101

(131)

.06

(1.00)

Long Task Duration (Block 3)

Low Trial Difficulty

Actiona 586

(142)

.45

(.80)

650

(196)

.59

(.96)

686

(196)

.64

(.85)

37

(127)

.05

(1.05)

Stateb 624

(150)

.31

(.48)

641

(125)

.50

(1.03)

685

(169)

.63

(.96)

44

(110)

.13

(1.31)

High Trial Difficulty

Actiona 620

(161)

.23

(.69)

632

(152)

.27

(.55)

690

(183)

.18

(.50)

58

(120)

-.09

(.68)

Stateb 657

(140)

.25

(58)

626

(126)

.50

(.89)

785

(215)

.88

(1.02)

159

(143)

.38

(.89)



We proceeded by subtracting participants’ mean response latencies on neutral trials from

participants’ mean response latencies on incongruent trials to obtain an index for Stroop

interference. A 2 (orientation) x 2 (trial difficulty) x 3 (task duration) ANOVA on Stroop

interference with repeated measures on the last two factors yielded a significant two-way interaction

between trial difficulty and task duration, F(2, 72) = 5.75, p < .006, ηp2 = .14. A closer examination

of this two-way interaction revealed a practice effect that was moderated by trial difficulty. As

predicted, Stroop interference on regular trials decreased as a function of task duration, F(2, 74) =

6.84, p < .003, ηp2 = .16, with greater Stroop interference in the first block (M = 140) compared to

the second block (M = 91) and the third block (M = 40). By contrast, there was no practice effect for

difficult trials, F < 1. Specifically, Stroop interference was similar in the first block (M = 107), the

second block (M = 80), and the third block (M = 101). Thus, as predicted, increased task difficulty

interfered with task practice effects.

The analysis also revealed the predicted three-way interaction effect, F(2, 72) = 5.22, p <

.009, ηp2 = .13. All relevant means are displayed in Table 4.2 (except for the means of the second

block, which were not affected by orientation and therefore, for the sake of brevity, not included in

the table). To facilitate interpretation of the three-way interaction, we analyzed the effects

separately for each block. Recall that the position of the blocks served as our operationalization of

task duration with the first block being less demanding than the second block, and the second block

being less demanding than the third block.

Block 1. A 2 (orientation) x 2 (trial difficulty) ANOVA yielded a significant main effect of

trial difficulty. Stroop interference was higher on regular trials than on difficult trials, F(1, 36) =

4.23, p < .05, ηp2 = .11 (M = 140 vs. M = 107). This main effect of trial difficulty was qualified by a

significant two-way interaction of orientation and trial difficulty, F(1, 36) = 6.00, p < .02, ηp2 = .14.

We conducted follow-up analyses to determine the locus of this unexpected effect. Action-oriented

participants displayed less Stroop interference on regular trials than state-oriented individuals, an


effect that was only marginally significant, F(1, 36) = 3.06, p = .09, ηp2 = .08 (M = 94 vs. M = 202).

By contrast, action- versus state-oriented participants displayed similar Stroop interference on

difficult trials, F < 1 (M = 110 vs. M = 101). Another way to interpret the results is to state that

state-oriented participants displayed more Stroop interference on regular trials than on difficult

trials, F(1, 15) = 6.97, p < .02, ηp2 = .32. By contrast, Stroop interference among action-oriented

individuals did not differ as a function of trial difficulty, F < 1.

Block 2. A 2 (orientation) x 2 (trial difficulty) ANOVA yielded no significant effects, all Fs

< 1. Specifically, action-oriented and state-oriented participants displayed similar Stroop

interference on difficult trials (M = 79 vs. M = 81) and on regular trials (M = 80 vs. M = 105).

Block 3. A 2 (orientation) x 2 (trial difficulty) ANOVA revealed a significant main effect of

trial difficulty, F(1, 36) = 7.82, p < .009, ηp2 = .18. Specifically, Stroop interference was higher on

difficult trials than on regular trials (M = 101 vs. M = 40). Furthermore, we found a non-significant

main effect of orientation, F(1, 36) = 2.07, p = .16, ηp2 = .05. More importantly, however, the

analysis revealed the expected two-way interaction between orientation and trial difficulty, F(1, 36)

= 5.56, p < .03, ηp2 = .13. Specifically, action-oriented and state-oriented participants displayed

similar Stroop interference on regular trials, F < 1 (M = 37 vs. M = 44). By contrast, action-oriented

individuals displayed lower Stroop interference than state-oriented individuals on difficult trials,

F(1, 36) = 6.15, p < .02, ηp2 = .15 (M = 58 vs. M = 159). Another way to interpret the two-way

interaction is to say that action-oriented participants showed similar Stroop interference on regular

and difficult trials, F < 1. By contrast, state-oriented individuals displayed more Stroop interference

on difficult trials than on regular trials, F(1, 15) = 9.19, p < .009, ηp2 = .38.

Errors. A 2 (orientation) x 2 (trial difficulty) x 3 (task duration) ANOVA on Stroop

interference in errors revealed no significant effects, all Fs < 1.6. Furthermore, there was no

indication for a speed/accuracy trade-off because the effects on Stroop interference in response

latencies remained intact when we controlled for Stroop interference in errors.


Letter task. Action-oriented participants’ error rates in the letter task were somewhat smaller

than state-oriented participants’ error rates in the letter task (M = 7% vs. M = 10%), F(1, 36) = 1.47,

p = .24, ηp2 = .04. Importantly, the effects on Stroop interference in response latencies remained

unchanged when we controlled for participants’ error rates in the letter task (Grand M = 8%).

Discussion

In Study 4.2, we varied level of demand by increasing the duration of the task and

manipulating task difficulty within the Stroop task. As expected, action- compared to state-oriented

participants displayed less Stroop interference on trials in the last of three consecutive blocks of

trials but not in the first or second block. Moreover, in line with our predictions, this effect emerged

on difficult trials (i.e., ink color evaluation followed by a letter evaluation) but not on regular trials

(i.e., solely ink color evaluation). Task practice in the third block likely decreased demands on

regular trials to an extent that action versus state orientation became irrelevant to efficiently

perform the regular Stroop trials. Notably, action- compared to state-oriented participants did not

generally perform better at difficult trials. The performance advantage of action-oriented

individuals only emerged after prolonged task duration made the task more demanding. Taken

together, Study 4.2 shows that the effects of action versus state orientation on cognitive control are

strongly moderated by the level of demand of the task context.

Stroop interference in Study 4.2 was not generally greater in the third block compared to the

first block. In fact, Stroop interference on regular trials declined as task duration proceeded,

whereas Stroop interference on difficult trials did not change. At first glance, this finding may seem

to contradict our assumption that sustained task engagement makes cognitive control tasks more

demanding. This apparent inconsistency can be resolved, however, if one considers that overt

cognitive control decrements across all participants are unlikely to occur within the comparatively

short task duration of Study 2 (which on average lasted 21 minutes). Indeed, Lorist and colleagues

(2005), for instance, report overt cognitive control decrements not until 30 minutes of continuous


task duration. To our knowledge, previous research did not consider the role of task difficulty,

which may have caused task practice to defer the occurrence of overt cognitive control decrements

during previous research. The present findings suggest, however, that task difficulty and action

versus state orientation are important factors to consider because they can moderate cognitive

control even at relatively short task durations.

Unexpectedly, participants’ performance in the first block revealed an interaction between

action versus state orientation and task difficulty. A closer examination revealed that the effects in

the first block were driven by decreased cognitive control among state-oriented participants on

regular compared to difficult trials. In the first block, task demands were still at a relatively low

level given that task duration had not yet progressed. Accordingly, state-oriented participants

showed no performance decrements on difficult trials. However, better performance on difficult

trials might have come at the cost of performance on regular trials. This would have been the case,

for instance, if state-oriented participants in the first block had formed a strong intention to perform

well on the following difficult trial. This intention may have undermined performance on the

intervening regular trials (cf. Kuhl & Helle, 1986; Shah et al., 2003). Although this explanation is

plausible, we did not explicitly predict the observed pattern in the first block. Future research is

therefore required to test the effects of shifting task priorities among action- versus state-oriented

individuals.

Study 4.3

In everyday life, people frequently encounter contexts that are more or less demanding.

During these real-life demands, people may have to plan and coordinate the pursuit of multiple

difficult and potentially incompatible goals. To the extent that such deliberations strain working

memory (Norman & Shallice, 1986), prolonged deliberation might lead to the depletion of

processing resources and less efficient use of working memory capacity (cf. Klein & Boals, 2001).

Because working memory is involved in cognitive control (Kane & Engle, 2003), real-life demands


may cause cognitive control decrements in much the same way as experimentally manipulated

demands. We designed Study 4.3 to examine this prediction.

Study 4.3 measured participants’ current level of demand in their personal lives. On the

basis of our theoretical analysis, we reasoned that the effects of real-life demands on cognitive

control would be moderated by action versus state orientation (cf. Baumann, Kaschel, & Kuhl,

2005). Among participants whose lives were currently demanding, we expected less cognitive

control in a Stroop task when participants were state-oriented as compared to when they were

action-oriented. By contrast, among participants whose lives were low-demanding, state-oriented

individuals should display similar or even greater cognitive control than action-oriented individuals

(cf. Jostmann & Koole, in press; Chapter 2; Koole et al., 2005).

Method

Participants and Design. Seventy-eight paid volunteers at the Vrije Universiteit Amsterdam

(29 men and 49 women; average age 20) participated in the experiment. The experimental design

was a 2 (orientation: action vs. state; between participants) x 2 (real-life demands: high vs. low;

between participants). Two participants failed a test for colorblindness and were therefore excluded

from the dataset. The main dependent variable consisted of participants’ average response latencies

and errors during the Stroop task.

Procedure. The equipment and general procedure were similar to Study 4.1. Participants

first answered a few questionnaires, including the ACS-90 (Kuhl, 1994b), to assess individual

differences in action versus state orientation (KR 20 = .64). Based on their responses, thirty-five

participants were assigned to the state-oriented group, whereas forty-one participants were assigned

to the action-oriented group. Participants then moved on with a Stroop task, which served as our

measure of cognitive control. Next, we assessed to what extent participants experienced their

current personal life as demanding. Finally, participants were tested for colorblindness, debriefed,

paid, and thanked by the experimenter.


Real-Life Demands. To assess real-life demands we asked participants to indicate to what

extent the following seven statements applied to them (1 = not at all; 7 = very much): Currently, I

am more busy than I usually am; Currently, I have particularly much to think of; I feel currently

under demand; I feel currently stressed out; I have a lot of work to do during the next days; I wish

my current life situation was more at ease; I find that others currently demand a lot from me.

Participants’ responses were averaged to obtain a single index of real-life demands (Cronbach’s

Alpha = .86). Participants who scored on average 4 or lower were assigned to the low demand

group (n = 37), whereas participants who scored on average higher than 4 were assigned to the high

demand group (n = 39).

Stroop task. The Stroop task in Study 4.3 was identical to the one in Study 4.1 with one

exception. We increased the number of trials such that there were 30 congruent trials (i.e., RED in

red, or BLUE in blue), 30 neutral trials (i.e., XXXX in red or blue), and 30 incongruent trials (i.e.,

RED in blue, or BLUE in red).

Results

Color Evaluation Latencies. Before analyzing response latencies, we removed erroneous

responses (3.0% of all responses) and responses faster than 300 ms (1.9% of all responses) from the

dataset. Initial analyses revealed that participants’ responses were quicker on neutral trials than on

incongruent trials, F(1, 75) = 46.43, p < .001, ηp2 = .38 (M = 524 vs. M = 581). Furthermore,

participants’ responses on congruent trials were as quick as responses on neutral trials, F < 1 (M =

524 vs. M = 524). We thus found interference but no facilitation.

We proceeded by subtracting average response latencies for neutral trials from average

response latencies for inconsistent trials to obtain an index of Stroop interference. A 2 (orientation)

x 2 (demand) ANOVA on Stroop interference revealed a significant interaction effect between

orientation and demands, F(1, 72) = 10.69, p < .003, ηp2 = .13. Relevant means are displayed in

Table 4.3. Among participants with high-demanding lives, those with an action orientation


displayed smaller Stroop interference effects than those with a state orientation, F(1, 72) = 4.06, p <

.05, ηp2 = .05 (M = 31 vs. M = 77). This pattern was reversed among participants with low-

demanding lives, such that those with an action orientation displayed higher Stroop interference

than those with a state orientation, F(1, 72) = 6.77, p < .02, ηp2 = .09 (M = 85 vs. M = 35).

Table 4.3 Response Latencies (in ms) and Errors in Color Evaluation as a Function of Orientation, Real-Life Demand, and Trial Type (Study 4.3; Standard Deviations Appear between Parentheses)

Trial Type


Participant Group Latency Errors Latency Errors Latency Errors Latency Errors

Low Real-Life Demand

Actiona 544

(109)

.45

(.51)

536

(106)

.55

(1.00)

622

(129)

1.85

(1.66)

85

(71)

1.30

(1.87)

Stateb 507

(94)

.88

(1.11)

513

(76)

.53

(.94)

547

(129)

1.24

(1.35)

35

(65)

.71

(1.45)

High Real-Life Demand

Actionc 506

(100)

.95

(1.86)

508

(83)

.81

(1.03)

539

(134)

1.33

(.46)

31

(68)

.52

(1.08)

Stated 539

(127)

.72

(1.07)

539

(128)

.55

(.78)

616

(189)

.72

(.75)

77

(88)

.17

(.86)

a n = 20. b n = 17. c n = 21. d n = 18. Note. SI = Stroop Interference (Incongruent – Neutral).

Another way to interpret the two-way interaction is to note that, among action-oriented

participants, those with high levels of demand displayed less Stroop interference than those with

low levels of demand, F(1, 72) = 10.35, p < .003, ηp2 = .13. By contrast, among state-oriented

participants, those with high levels of demand displayed non-significantly more Stroop interference

that those with low levels of demand, F(1, 72) = 2.19, p = .15, ηp2 = .03.

Errors. A 2 (orientation) x 2 (demand) ANOVA on Stroop interference in errors revealed a

significant main effect of demand, F(1, 72) = 4.34, p < .05, ηp2 = .06. Participants with low-


demanding lives displayed higher Stroop interference in errors than participants with demanding

lives (M = 1.03 vs. M = .36). There was no main effect of orientation, F(1, 72) = 2.27, p = .14, ηp2 =

.03, and no interaction effect, F < 1. Given that the main effect of demand was not anticipated, we

conducted a (orientation) x (demand) ANOVA on Stroop interference in response latencies with

Stroop interference in errors as a covariate, which left the results intact. Furthermore, Stroop

interference in response latencies and Stroop interference in errors were uncorrelated (r = .06, ns).

Thus, lower Stroop interference in response latencies was not due to a speed/accuracy trade-off.

Discussion

As predicted, participants who reported having demanding lives displayed greater cognitive

control when they were action- rather than state-oriented. The results of Study 4.3 thus provide a

conceptual replication of Study 4.2 using real-life demands rather than task duration to

operationalize demands. By contrast, participants who reported having low-demanding lives

displayed greater cognitive control when they were state- rather than action-oriented. The latter

finding fits with the assumption that, under low-demanding conditions, state-oriented individuals

can function equally or even more effectively than action-oriented individuals (Jostmann & Koole,

in press; Chapter 2; see also Koole et al., 2005). An additional finding of Study 4.3 was that action-

oriented participants displayed greater cognitive control when demands were high rather than low.

This finding indicates that action-oriented individuals’ performance may actually benefit from

increased levels of demand.

Study 4.4

Why would high demands induce enhanced cognitive control among action-rather than

state-oriented individuals? In Study 4.4, we examined whether this pattern might be attributed to a

goal neglect mechanism. To overcome goal neglect, a person has to correctly remember the

intention that has to be executed and, importantly, initiate cognitive control in the appropriate

context. By contrast, the ability to overcome goal neglect is irrelevant if a person merely has to


maintain cognitive control. Once cognitive control is successfully initiated, executive systems

remain in a high-control state (Botvinick et al., 2001), which facilitates subsequent efforts to exert

cognitive control. Taken together, a goal neglect account predicts that action versus state orientation

moderates the impact of demands on cognitive control when cognitive control has to be initiated but

not when cognitive control merely has to be maintained.

Study 4.4 examined goal neglect processes by manipulating the availability of prompts (e.g.

De Jong et al., 1999; Dibbelt, 1997; Kane & Engle, 2003). When prompts are available, the

initiation of cognitive control no longer relies on the internal activation of the task intention.

Accordingly, prompts make cognitive control less dependent on goal neglect. By contrast, when no

prompts are available to initiate cognitive control, it is necessary to keep the task intention

continuously accessible in working memory. Accordingly, the absence of prompts makes cognitive

control more susceptible to goal neglect. Past research on Stroop interference has varied the

composition of Stroop trials to manipulate the availability of prompts (Kane & Engle, 2003). To

connect our findings to this work, we decided to manipulate prompts in a similar way.

In a Stroop task that was adapted from Kane and Engle (2003), we embedded critical trials

in a context of filler trials. Filler trials were either congruent (75% congruent condition) or

incongruent (12,5% congruent condition). Following Kane and Engle, we reasoned that a context of

congruent filler trials encourages participants to disregard the task instruction to ignore the word

meaning. Accordingly, every time participants encountered one of the rare incongruent trials in the

context of mainly congruent trials they had to re-initiate cognitive control. The context of congruent

trials thus made cognitive control highly sensitive to goal neglect. By contrast, a context of

incongruent filler trials continuously prompts the task intention to ignore the word meaning.

Accordingly, participants in the 12,5% congruent condition only needed to maintain cognitive

control. A context of incongruent filler trials thus made cognitive control less sensitive to goal

neglect. Theoretically, we regard goal neglect as a key mechanism that explains the functional


differences between action- and state-oriented individuals. Therefore, we expected high demands to

evoke performance differences between action- and state-oriented participants in the 75% congruent

condition but not necessarily in the 12,5% congruent condition.

The mechanism that underlies cognitive control has important implications for whether the

effects of action versus state orientation and high demands should be expected on response latencies

versus errors. As noted in the Introduction, cognitive control deficits on incongruent trials in a

context of predominantly congruent trials (75% congruent condition) are more likely to be reflected

by increased errors rather than an overall increase in response time latencies (Kane & Engle, 2003).

By contrast, when preceding trials are predominantly incongruent (12,5% congruent condition),

performance deficits on incongruent trials are more likely to be manifest in a shift towards slower

response time latencies rather than increased errors. Because we assumed that action versus state

orientation moderates the vulnerability to goal neglect under high demands, we expected the effects

in Study 4.4 to emerge in errors rather than response latencies.

To induce varying levels of demand, we induced a prospective memory load among half of

the participants. Specifically, we told participants in the high-demanding condition that they would

have to perform a task at the end of the experimental session for which they could earn an extra

financial reward. We reasoned that this instruction would strain working memory (Goschke &

Kuhl, 1993; Jostmann & Koole, in press; Chapter 2; Kuhl & Helle, 1986) and thereby give rise to a

demanding context. Participants in the low-demanding condition did not receive any hints about an

upcoming task. We expected that the prospective memory load would trigger different mobilization

of control resources among action- versus state-oriented participants, such that prospective memory

load lowers Stroop interference among action-oriented compared to state-oriented participants.


Method


Participants were 155 paid volunteers at the Vrije Universiteit Amsterdam (52 men and 103

women, average age 20). The experimental design was 2 (orientation: action vs. state; between

participants) x 2 (prospective load: high vs. low; between participants) x 2 (congruency proportion:

75% congruent vs. 12,5% congruent, between participants). Fourteen participants failed to pass a

colorblindness test and were therefore removed from the dataset. The main dependent variables

consisted of participants’ average response latencies and errors in a Stroop task.

Procedure

The general procedure and the equipment were similar to Studies 4.1 and 4.3. Participants

were randomly assigned to one of the four experimental conditions (prospective load – 75%

congruent: n = 35; prospective load – 12,5% congruent: n = 37; no prospective load – 75%

congruent: n = 34; no prospective load – 12,5% congruent: n = 35). Participants started by filling

out some questionnaires, which contained our assessment of individual differences in action versus

state orientation (KR 20 = .77). Based on their responses, eighty-four participants were classified as

state-oriented, and seventy-one participants were classified as action-oriented.4.5 Participants moved

on with a study that contained our manipulation of prospective memory load. After this, participants

continued with the Stroop task, which served as our measure of cognitive control. Next, participants

were tested for colorblindness, after they were paid, debriefed, and thanked for their participation.

Prospective Memory Load Manipulation. After the individual differences assessment, we

manipulated prospective memory load. The procedure, which was developed by Koole and

Jostmann (2004), required participants to first solve a number of simple arithmetic sums. During

each sum, participants had to add up three one- or two-digit numbers and type in the correct

response. After participants had solved one practice sum, they were given 2.5 minutes to solve as

many sums as possible. During the entire task, the remaining time was displayed on the computer


screen. After 2.5 minutes had passed, participants were given feedback regarding the number of

sums they had solved correctly. Subsequent to the feedback, the prospective load was induced.

Participants in the prospective load condition were told that there would be a second block

of arithmetic sums. Participants were further informed that the level of difficulty and amount of

time for the second block were identical to the first block. However, during the second block,

participants could earn a financial bonus if they improved their performance compared to the first

block. Specifically, participants would receive a bonus of €1.00 if they solved at least 10% more

sums and a bonus of €2.50 if they solved at least 25% more sums during the second block. It was

further explained that the second block of arithmetic sums would be preceded by several unrelated

studies, ostensibly to avoid fatigue. These unrelated studies included our measure of cognitive

control. Participants in the no prospective load condition were told that they had solved more sums

during the first block than the average participant within the same amount of time. Therefore, they

received an extra bonus of €2.50. The unexpected reward was provided to keep the total financial

compensation constant across conditions. In the no prospective load condition, no mention was

made of a second block of arithmetic sums. All participants completed a second block of arithmetic

sums after the Stroop task.

Stroop task

The Stroop task, which was based on Kane and Engle (2003), was similar to Study 4.1 with

the following exceptions: Specifically, the task consisted of 160 trials, 60 of which were critical for

analysis. The remaining 100 trials were filler trials, which, however, were not identifiable by the

participants as fillers. Trials were randomly defined as respectively filler trials or critical trials

before the experiment was conducted. All analyses reported in the Results section include only

critical trials. (Analyses including critical trials and filler trials revealed equivalent results.)

Participants saw 20 critical trials of each trial type (incongruent, neutral, congruent). In the 12,5%

congruent condition, filler trials were incongruent. Participants in the 12,5% condition thus saw 20


congruent trials, 20 neutral trials, and 120 incongruent trials. By contrast, in the 75% congruent

condition, filler trials were congruent. Accordingly, participants in the 75% congruent condition

saw 120 congruent trials, 20 neutral trials, and 20 incongruent trials. All color words and ink colors

were equally represented across filler and critical trials (except for the neutral letter string XXXX,

which was only presented during critical trials.) The presentation of all 160 trials varied randomly

for each participant.

Results

Manipulation Checks. After participants had completed the second block of arithmetic sums,

we asked participants to indicate why they would receive an extra bonus. One hundred thirty-seven

participants (96.5% of the entire sample) provided the correct answer. Rerunning the main analyses

without the five participants who failed this test (3.5% of the entire sample) yielded no

substantively different results, so we retained them in the dataset. We further assessed how

participants currently felt (1 = very bad, 9 = very good) before the prospective load manipulation,

and how they felt directly after the manipulation but before the Stroop task. To obtain an index of

mood change as a function of prospective load, we subtracted participants’ mood ratings before the

manipulation from their mood ratings after the manipulation with higher scores indicating greater

mood improvement (Grand M = .49). A 2 (orientation) x 2 (prospective load) ANOVA revealed

greater mood improvements when prospective load was absent rather than when it was present (M =

.90 vs. M = .10), F(1, 137) = 6.76, p < .02, ηp2 = .05. No effects involving orientation occurred, all

Fs < 1. Mood improvements were not related to Stroop interference in response latencies (r = -.01,

ns) or to Stroop interference in errors (r = -.02, ns).


Table 4.4 Response Latencies (in ms) and Errors in Color Naming as a Function of Orientation, Congruency Proportion, Prospective Memory Load, and Trial Type (Study 4.4; Standard Deviations Appear between Parentheses)

Trial Type Congruent Neutral Incongruent SI

Participant Group

Latency Errors Latency Errors Latency Errors Latency Errors

12,5% Congruent

No Prospective Load

Actiona 515

(118)

.20

(.41)

538

(118)

.50

(.69)

539

(123)

.60

(.94)

1

(48)

.10

(.85)

Stateb 476

(101)

.20

(.41)

496

(116)

.13

(.35)

530

(116)

.20

(.41)

35

(41)

.07

(.59)

Prospective Load

Actionc 498

(85)

.28

(.58)

504

(93)

.39

(.50)

536

(130)

.89

(1.18)

31

(64)

.50

(.99)

Stated 503

(71)

.21

(.54)

492

(61)

.32

(.58)

550

(95)

.37

(.60)

58

(59)

.05

(.62)

75% Congruent

No Prospective Load

Actione 549

(125)

.07

(.26)

597

(133)

.20

(.41)

712

(230)

1.53

(1.85)

114

(155)

1.33

(1.63)

Statef 536

(163)

.37

(.68)

567

(204)

.32

(.48)

678

(313)

.90

(1.37)

111

(155)

.58

(1.43)

Prospective Load

Actiong 483

(113)

.67

(.82)

536

(140)

.40

(.74)

591

(242)

1.20

(1.21)

56

(177)

.80

(1.42)

Stateh 490

(94)

.35

(.49)

533

(119)

.30

(.66)

634

(205)

2.10

(1.86)

102

(121)

1.8

(1.91)

a n = 20. b n = 15 c n = 18. d n = 19. e n = 15. f n = 19. g n = 15. h n = 20. Note. SI = Stroop Interference (Incongruent – Neutral).


Stroop Effect

Color Evaluation Latencies. Before analyzing response time latencies, we removed

erroneous responses (2.7% of all responses) and responses faster than 300 ms (4.6% of all

responses) from the dataset. On average, participants’ responses on neutral trials were faster than

responses on incongruent trials, F(1, 140) = 46.24, p < .001, ηp2 = .25 (M = 532 vs. M = 594). In

addition, average responses on congruent trials were faster than responses on neutral trials, F(1,

140) = 36.36, p < .001, ηp2 = .21 (M = 506 vs. M = 532). Thus, color evaluation latencies revealed

evidence of both interference and facilitation. Next, we subtracted participants’ mean response

latencies on neutral trials from participants’ mean response latencies on incongruent trials to obtain

an index for Stroop interference. A 2 (orientation: action vs. state) x 2 (prospective load: high vs.

low) x 2 (congruency proportion: 75% congruent vs. 12,5% congruent) ANOVA on Stroop

interference yielded a significant main effect of congruency proportion, F(1, 133) = 7.77, p < .007,

ηp2 = .06. There was less Stroop interference in the 12,5% congruent condition than in the 75%

congruent condition (M = 30 vs. M = 97). The effect of congruency proportion was qualified by

orientation nor prospective load, all ps > .11.

Errors. Initial analysis revealed that participants conducted more errors on incongruent trials

than on neutral trials, F(1, 140) = 31.60, p < .001, ηp2 = .18 (M = .98 vs. M = .33). Participants’

errors on neutral trials did not differ from those on congruent trials, F < 1 (M = .33 vs. M = .29).

Thus, error data revealed evidence of interference but no facilitation.

Next, we subtracted participants’ mean errors on neutral trials from participants’ mean errors

on incongruent trials to obtain an index for Stroop interference. We then analyzed Stroop

interference using a 2 (orientation) x 2 (prospective load) x 2 (congruency proportion) ANOVA,

which revealed a main effect of congruency proportion, F(1, 133) = 19.43, p < .001, ηp2 = .13. On

average, Stroop interference was lower in the 12,5% congruent condition than than in the 75%

congruent condition (M = .18 vs. M = 1.15). However, this main effect was qualified by the


predicted three-way interaction between orientation, prospective load, and congruency proportion,

F(1, 133) = 6.35, MSE = 1.607, p < .02, ηp2 = .05. Relevant means are displayed in Table 4.4.

We proceeded with separate analyses for the 75% congruent condition and the 12,5%

congruent condition. Recall that we hypothesized that effects of orientation and prospective load on

cognitive control were caused by goal neglect. Accordingly, effects of orientation should especially

be apparent when prompts were absent, that is, in the 75% congruent condition. Consistent with

this, a 2 (orientation) x 2 (prospective load) ANOVA found no significant effects in the 12,5%

congruent condition, all ps > .2. By contrast, the same 2 (orientation) x 2 (prospective load)

ANOVA in the 75% congruent condition yielded a significant two-way interaction between

orientation and prospective load, F(1, 65) = 4.94, p < .04, ηp2 = .07. When prospective load was

present, state-oriented participants displayed greater Stroop interference than action-oriented

participants, F(1, 133) = 5.33, p < .03, ηp2 = .04 (M = 1.80 vs. M = .80). However, when

prospective load was absent, the pattern was reversed. State-oriented participants showed a

marginally smaller Stroop interference effect than action-oriented participants, F(1, 133) = 2.97, p =

.087, ηp2 = .04 (M = .58 vs. M = 1.33).

Another way of interpreting the orientation by prospective load interaction in the 75%

congruent condition is to note that, among state-oriented participants, Stroop interference was

higher when prospective load was present rather than absent, F(1, 133) = 9.04, p < .004, ηp2 = .06.

Among action-oriented participants, however, we found a non-significant reversal, such that Stroop

interference was greater when prospective load was absent rather than present, F(1, 133) = 1.33, p =

.26, ηp2 = .01.

Supplementary Analyses. Effects on Stroop interference in errors remained unchanged when

we controlled for Stroop interference in response latencies. Thus, there was no indication for a

speed/accuracy trade-off. Moreover, effects on Stroop interference in errors remained unchanged

when we controlled for the number of solved sums during the first arithmetic task (Grand M =


14.2), or for mood improvements induced by the prospective load manipulation (Grand M = .49).

Finally, for each participant, we computed an index of performance improvement by subtracting the

number of correctly solved arithmetic sums during the first block from the number of correctly

solved sums during the second block. Action- versus state-oriented participants displayed similar

performance improvements during the second block of arithmetic sums (Grand M = .7). Stroop

interference in errors was not affected by performance improvements.

Discussion

The results of Study 4.4 illuminate some of the mechanisms through which action versus

state orientation moderates the impact of demands on cognitive control. In line with a goal neglect

account, action versus state orientation moderated cognitive control when no prompts for cognitive

control were available in the 75% congruent condition. In addition, the effects of action versus state

orientation emerged on errors rather than response latencies. Previous work suggests that, in the

context of the paradigm used in Study 4.4, errors are the most sensitive index of goal neglect

processes (Kane & Engle, 2003). Taken together, goal neglect appears to be a critical mechanism

that explains why action versus state orientation moderates the impact of demands on cognitive

control (cf. Jostmann & Koole, in press).

When the task context discouraged goal neglect by providing prompts (i.e., in the 12,5%

congruent condition), neither errors nor response latencies revealed any differences in cognitive

control as a function of demands and action versus state orientation. Previous work indicates that

the availability of prompts in the paradigm used in Study 4.4 does not remove the competition

between the competing ink color and word color dimensions (Kane & Engle, 2003). Accordingly,

the basic capacity for competition resolution seems to remain unaffected by action versus state

orientation (cf. Baumann & Kuhl, 2005; Dibbelt, 1997; Fuhrmann & Kuhl, 1998). However, this

conclusion may need to be qualified in light of the potential occurrence of sequence or practice

effects during repeated response activations (cf. Mayr, Awh, & Laurey, 2003). To the extent that


sequence or practice effects have reduced the required amount of cognitive control below a critical

level in the 12,5% congruent condition, participants' action versus state orientation may have

become irrelevant. Future research may address this possibility by including a trial type that is more

resistant against practice effects (see the difficult trials in Study 4.2), or by increasing the number of

different ink colors to reduce the likelihood of response repetitions (cf. Kerns, Cohen, MacDonald,

Cho, Stenger, & Carter, 2004).

In the low-demanding condition in Study 4.4, participants experienced a significant mood

improvement. Accordingly, one might wonder whether the effects in Study 4.4 could be

alternatively explained by mood improvements in the low-demanding condition rather than, as we

suggest, by prospective load in the high-demanding condition. Supplementary analyses render a

mood-based explanation less likely because mood improvements were not related to Stroop

interference. In addition, all analyses on Stroop interference remained intact when we statistically

controlled for mood improvements. Finally, unlike Stroop interference, mood improvements after

the demand induction were not affected by action versus state orientation (cf. Koole & Jostmann,

2004, Experiment 1). We therefore conclude that the effects in Study 4.4 were unlikely to result

from mood improvements in the low-demanding condition.

General Discussion

In the present research, we investigated how action versus state orientation regulates the

impact of high demands on cognitive control. Across four studies, we found evidence that high

demand leads to greater cognitive control in Stroop tasks among action- compared to state-oriented

individuals. By contrast, low demand led action-oriented individuals to display comparable or lower

levels of cognitive control compared to state-oriented individuals. Our findings were consistent

across different manipulations of demands. Action- compared to state-oriented participants

displayed greater cognitive control after the performance of a demanding task (Studies 4.1 and 4.2),

and under prospective load (Study 4.4). Notably, action versus state orientation moderated cognitive


control even in response to real-life demands (Study 4.3). Taken together, action versus state

orientation moderates cognitive control across different operationalizations of demand.

We identified goal neglect as a key mechanism that explains the functional differences

between action-oriented and state-oriented individuals under high demand. Specifically, high

demand in Study 4.4 led action-oriented participants to display better cognitive control compared to

state-oriented participants when task conditions were conducive to goal neglect. Because goal

neglect reflects an insufficiency of working memory during cognitive control (Kane & Engle,

2003), our findings suggest that improved cognitive control under high demands is due to more

efficient use of working memory among action- compared to state-oriented individuals. This idea is

consistent with recent findings that under high demands action-oriented individuals utilize their

working memory capacity more efficiently than state-oriented individuals as indicated by

individuals’ OSPAN scores (Jostmann & Koole, in press; Chapter 2). Taken together, action-

oriented individuals seem to be better protected against goal neglect under high demand than state-

oriented participants.

The conditions that trigger action versus state orientation to regulate cognitive control

theoretically involve sustained working memory use. Such sustained use may be caused by

prolonged engagement in difficult tasks or persisting prospective load. Action versus state

orientation therefore should not be triggered by task difficulty per se but rather by the draining

effects of prolonged task difficulty. In line with this reasoning, action-oriented participants did not

outperform state-oriented participants on the difficult working memory task at the beginning of

Study 4.1 but only on the subsequent Stroop task. Moreover, we found no performance differences

in Study 4.2 between action- and state-oriented participants on difficult trials in the first two blocks

but only in the third block after task engagement had been progressed. In short, action versus state

orientation regulates cognitive control when difficult task conditions persist over time. The

importance of prolonged task difficulties may be related to the activity of dopaminergic circuits in


the brain, which become depleted after prolonged activation of working memory (cf. Lorist et al.,

2005).

The effects of action versus state orientation have some similarities with the effects of

working memory capacity. Both constructs moderate cognitive control, especially under conditions

that encourage goal neglect (Kane & Engle, 2003). However, there also exist important differences.

Unlike individuals high in working memory capacity, action-oriented individuals displayed

cognitive control improvements only under high-demanding circumstances. Under low-demanding

circumstances, action-oriented individuals performed equally well (Studies 4.2 and 4.4) and

sometimes even worse than state-oriented individuals (Study 4.3; cf. Koole et al., 2005). In

addition, action versus state orientation did not moderate cognitive control under conditions that

discourage goal neglect (Study 4.4). By contrast, such moderating effects have been reported for

working memory capacity under almost identical task conditions, although the effects were not

always robust across experiments (Kane & Engle, 2003). Finally, action versus state orientation

moderated cognitive control over and above individual differences in working memory capacity

(Study 4.1). In sum, although there exist some similarities between the two constructs, the effects of

action versus state orientation on cognitive control are conceptually and empirically distinct from

the effects of working memory capacity.

Based on the present research and recent findings (Jostmann & Koole, in press; Chapter 2;

cf. Kuhl, 2000), we suggest that individual differences in working memory capacity reflect general

limitations in cognitive control, whereas action versus state orientation reflects the ability to use

one’s working memory capacity efficiently to exert cognitive control under high demands.

Importantly, higher working memory efficiency makes it unnecessary for action-oriented

individuals to mobilize extra efforts (cf. Brehm & Self, 1989; Eysenck & Calvo, 1992). Under high

demands, increased effort mobilization during cognitive control may lead to subsidiary costs as

indicated by a speed/accuracy trade-off (Boksem, 2006; cf. Hockey, 1997). No such speed/accuracy


trade-off was found in the present research. It thus appears that action versus state orientation under

demands relates to differences in processing efficiency rather than capacity, or effort mobilization.

Limitations and Future Directions

The present research remains preliminary and therefore leaves many questions open for

further investigation. One important task for future research is to gauge the scope of the impact of

action versus state orientation on executive functions. One recent analysis suggests the separability

of three executive functions including mental set shifting, information updating, and inhibition of

prepotent responses (Miyake et al., 2000). The present research on cognitive control in Stroop tasks

has shown that action versus state orientation moderates the impact of high demands on the third set

of functions. Moreover, initial evidence from our laboratory suggests that action versus state

orientation also moderates the effects of high demands on more complex executive tasks including

the OSPAN task (Jostmann & Koole, in press; Chapter 2), or the Tower-of-Hanoi task (Jostmann,

Gieselmann, & Koole, 2004). Future research is needed to examine whether the effects of action

versus state orientation also apply to other executive functions such as set shifting and updating.

The present findings might also be relevant for more applied research. The present research

revealed that state-oriented participants under high demands display some symptoms that are

characteristic of depression, such as lack in initiative (Hertel & Hardin, 1990) or “forgetting” to

enact one’s intentions (Kuhl & Helle, 1986). Although the present research focused exclusively on

non-clinical populations, PSI theory (Kuhl, 2000; Kuhl & Beckmann, 1994a) suggests a link

between state orientation and vulnerability to depression. Future research on the antecedents of

depression might therefore benefit from considering the role of state orientation and chronically

elevated levels of demand. Moreover, the relative performance decrements among state-oriented

individuals under high demands raise the question how working, learning and testing conditions

should be structured in order to provide ideal environments for employees or students with an

action versus state orientation, respectively (cf. Diefendorff et al., 1998, 2000). In sum, applied


research might profit from considering the role of action versus state orientation under high

demands.

Concluding Remarks

The capacity to exert cognitive control is a fundamental prerequisite for intentional action.

The present research suggests that at least some people may be capable of strategically adjusting

their capacity for cognitive control to the demands of the situation. Indeed, we found consistent

evidence that action-oriented individuals strengthen their cognitive control under high-demanding

circumstances in comparison with state-oriented individuals. As such, action orientation may be a

vital resource in the regulation of cognitive control under demanding conditions.


Chapter 5

Dealing with Demands

Conclusions, Integration, and Implications

The present dissertation highlights the role of action control in coping with high demands. In

Chapter 1, I defined action control as the mental processes that allow people to form, maintain, and

implement (or disengage from) their intentions. Action control presumes effective management of

working memory. High demands involve sustained activation of working memory, which reduces

the effectiveness of this system. I therefore reasoned that effective action control should moderate

the psychological impacts of high demands. Accordingly, action-oriented individuals, who are

highly skilled at action control, may shield themselves more effectively against high demands than

state-oriented individuals, who are less skilled at action control.

Because action control operates at a global level of psychological functioning, action control

should moderate the impact of high demands across many major domains of human functioning,

including cognition, affect, and behavior. Chapter 1 reviewed previous research that provided some

suggestive support for this notion. This previous research suffered from important limitations, and

no research systematically examined how action versus state orientation moderates the impact of

high demands on cognition, affect, and behavior. I therefore conducted a program of research aimed

at clarifying the role of action versus state orientation in cognitive, affective, and behavioral

shielding against high demands. This research is described in Chapters 2 to 4.

In this final chapter, I summarize the main theoretical conclusions that can be drawn from

Chapters 2 to 4. In addition, I make an initial attempt to integrate the findings that emerged from the

present dissertation research into a coherent theoretical model. This model, which is based on recent

models of intentional action (Kuhl, 2000) and working memory (Braver & Cohen, 2000; Cohen,

Braver, & Brown, 2002), attributes a central role to the updating function of working memory.


Action Orientation and Cognitive, Affective, and Behavioral Shielding

Cognitive shielding refers to the extent to which people are able to protect their working

memory resources from the burden of demands. Chapter 2 found evidence that action- compared to

state-oriented individuals display improved cognitive performance under high demands on tasks

that rely on working memory. High demands were induced by having participants visualize a

demanding relationship, whereas low demands were induced by having participants visualize an

accepting relationship (Baldwin & Sinclair, 1996). Visualizing a demanding relationship led to

higher operation span (OSPAN) scores (Study 2.1; cf. Turner & Engle, 1989) and greater intention

memory in a postponed intention task (Study 2.2; cf. Goschke & Kuhl, 1993) than visualizing an

accepting relationship among a action-oriented individuals. By contrast, state-oriented individuals

had lower OSPANs (Study 2.1) and smaller intention memory (Study 2.2) after visualizing a

demanding compared to an accepting relationship. In sum, action- but not state-oriented individuals

diplay improved cognitive shielding under high demands.

Affective shielding refers to the extent to which people are able to maintain or restore a

positive affective state, particularly under demanding conditions. In Chapter 3, I found evidence

that action-oriented individuals are also more effective at affective shielding than state-oriented

individuals (cf. Koole & Jostmann, 2004). Using a subliminal priming task (Chartrand, Van Baaren

& Bargh, 2006), participants in Study 3.1 were exposed to schematic faces with either an angry,

neutral, or happy expression. Among participants who were primed with angry faces, those with an

action orientation showed a more positive affective reaction than those with a state orientation on a

basic affect measure (Tesser, Millar, & Moore, 1988). No similar affective differences between

action- and state-oriented individuals were found among participants who were primed with neutral

or positive faces. Moreover, state-oriented individuals displayed a more negative affective reaction

when they were primed with angry compared to happy faces, whereas no such effect was found

among action-oriented individuals. Taken together, Chapter 3 found evidence that action- compared


to state-oriented individuals display more efficient affective shielding, even when affect is

processed at unconscious levels.

Behavioral shielding refers to the extent to which people are able to enact their intentions.

Chapter 4 operationalized behavioral shielding as participants’ ability to exert cognitive control in

Stroop tasks, which required participants to override a strong but task inappropriate response

tendency. High demands were manipulated by having participants perform a difficult task that

strains working memory (Studies 4.1 – 4.2; cf. Kazén & Kuhl, 2005; Turner & Engle, 1989), or by

inducing a norm for future task performance (Studies 4.4; cf. Koole & Jostmann, 2004). Moreover,

Chapter 4 also measured the level of demand in participants’ personal lives (Study 4.3; cf.

Baumann, Kaschel, & Kuhl, 2005). Across these different demand manipulations, high demands led

to greater cognitive control among action- compared to state-oriented individuals. No similar

differences in cognitive control were found under low demands. Chapter 4 thus found evidence that

action- compared to state-oriented individuals are more effective at behavioral shielding under

demanding conditions.

Chapter 4 further shed light on one mechanism whereby action orientation facilitates

behavioral shielding under high demands. More specifically, improved behavioral shielding under

high demands among action- compared to state-oriented individuals was found to depend on

whether behavioral shielding requires the active representation of an intention in working memory

(Study 4.4). When the task intention to ignore the word meaning during a Stroop trial was prompted

by a large proportion of incongruent Stroop trials (cf. Kane & Engle, 2003), action- versus state-

oriented individuals displayed no differences in cognitive control in response to varying levels of

demand. By contrast, when participants had to maintain the task intention in working memory

because no external prompts were available, high demands led to increased cognitive control among

action- compared to state-oriented individuals. Taken together, the findings of Study 4.4 suggest


that behavioral shielding among action- versus state-oriented individuals under high demands

depends on whether people can utilize their working memory efficiently to guide cognitive control.

An Integrative Model of Action Control and Coping with Demands

The central theme of the present dissertation is that action-oriented individuals shield

themselves better than state-oriented individuals against the psychological costs of high demands in

cognition, affect, and behavior. In psychology, cognition, affect, and behavior are widely

recognized as the main components of psychological functioning. It thus appears that action versus

state orientation thus regulates the psychological impact of demands across all main domains of

human functioning. This very general influence of one variable might be considered surprising,

because many psychologists today focus on processes and phenomena that are highly specific to

particular paradigms and experimental tasks. How can a single variable like action versus state

orientation predict such broad aspects of human functioning?

One explanation for the observed generality across domains is that the different types of

shielding are related to another because they each contribute to intentional action control.

Specifically, cognitive shielding facilitates the formation and maintenance of intentions. Moreover,

affective shielding processes block negative mood and the activation of intention-inappropriate

action tendencies. Finally, behavioral shielding facilitates the implementation of intentions by

overriding competing impulses. Because overall action control can only be effective to the extent

that all subcomponents function effectively, it is likely that improved shielding in one domain may

also lead to improved shielding in other domains. In the following, I will further elaborate on this

idea by proposing an integrative model of action control and coping with high demands.


Figure 5.1. A Functional Model of Intentional Action Control

(Adapted from PSI theory; Kuhl, 2000; Concepts in Italics are Added by the Author).

This theoretical integration is based on a recent model of intentional action control that has

been elaborated in personality systems interactions (PSI) theory (Kuhl, 1984, 2000, 2001).

According to PSI theory, effective action control depends on the interaction between high-level

mental systems like working memory (and in particular intention memory; cf. Kazén & Kuhl, 2005)

and low-level behavioral output systems (see Figure 5.1). When the neuro-cognitive pathway

between working memory and behavioral output systems is inhibited, no information between these

two types of systems can be exchanged. In such cases, people’s behavior is likely to be determined

by automatic behavior routines. However, when the pathway is facilitated, intentions that are

represented in working memory can be released to make them accessible for implementation

through behavioral output systems. Moreover, pathway facilitation also allows perceptual

information to enter working memory thereby changing its contents. Such perceptual information

may include sensory information but also information about personal goals and needs. Taken

together, pathway facilitation leads to what recent neuropsychological models (Braver & Cohen,

2000; Cohen et al., 2002) would refer to as the updating of working memory.

PSI theory (Kuhl, 2000, 2001) further proposes that the pathway between working memory

and behavioral output systems is mediated by positive affect. Specifically, high positive affect

Behavioral Output Systems

Dopamine/ Positive Affect

Working Memory

(Intention Memory)


facilitates the pathway, whereas low positive affect leads to its inhibition (for experimental

evidence, see Kazén & Kuhl, 2005; Kuhl & Kazén, 1999). On a neurological level, high positive

affect has been associated with increased activity of the neurotransmitter dopamine (Ashby, Isen, &

Turken, 1999). Consistent with the predictions of PSI theory regarding the role of positive affect,

recent neuropsychological models (Braver & Cohen, 2000; Cohen et al., 2002) suggest that

increased dopaminergic activity in relevant brain regions (e.g., prefrontal cortex) facilitates the

updating of working memory. Recent research indeed suggest that the experimental induction of

high positive affect, and a genetic disposition for high dopaminergic activity both facilitate

intention implementation in a cognitive control task on the one hand, and the uptake of new

information in working memory on the other hand (Dreisbach et al. 2005; Dreisbach & Goschke,

2004; cf. Olivers & Nieuwenhuis, 2006). Taken together, high positive affect or high dopaminergic

activity facilitate the updating of working memory.

Based on the foregoing analysis, I suggest that successful action control depends on how

well people can update their working memory. This updating mechanism, in turn, may explain the

threefold effect on action control including cognitive, affective, and behavioral effects. Specifically,

updating may contribute to cognitive shielding because, on the one hand, it causes old and

potentially irrelevant information to be erased and, on the other hand, it allows new and potentially

more relevant information to enter working memory.5.1 Moreover, because updating is related to

increased positive affect, updating may also help people to shield against negative affect. Finally,

updating may contribute to behavioral shielding because it makes intentions that are represented in

working memory, accessible to guide behavioral output. If this line of reasoning is correct, the

effects of high demands on the shielding of cognition, affect, and behavior may be mediated by the

updating function of working memory.

Under high demands, the updating function likely becomes temporarily impaired. The

reason is that high demands involve the sustained activation of working memory, which, in turn,


leads to declines in dopaminergic activity. In line with this idea, recent research suggests that

cognitive impairments after prolonged engagement in a cognitive control task are caused by

decreases in dopaminergic activity (Lorist, Boksem, & Ridderinkhof, 2005). To the extent that

working memory updating is involved in cognitive control (cf. De Jong, Berendsen, & Cools, 1999;

Kane & Engle, 2003), the findings of this research thus support the idea that high demands impair

the updating function of working memory. Consequently, an effective way to cope with high

demands is thus to maintain the updating function of working memory. Because the updating

function is an essential part of action control, high skills at action control among action-oriented

individuals lead to more effective coping with high demands than low action control skills among

state-oriented individuals.

In sum, the “updating model” of action control and coping with high demands holds that

effective action control involves the updating function of working memory. Efficient updating of

working memory contributes to triple shielding of cognition, affect, and behavior. Updating is

facilitated by increased positive affect and high activity of its neurological substrate dopamine.

Under high demands, dopaminergic activity declines thereby leading to a transient impairment of

the updating function. Coping with high demands requires people to restore the updating function.

Accordingly, the updating model predicts that action-oriented individuals cope better with high

demands than state-oriented individuals because they can regulate the updating function more

effectively under high demands than state-oriented individuals.

The updating model integrates the findings of the present dissertation and links them to

contemporary discussions of action control and working memory. Although the ultimate validity of

the model remains to be tested by future research, the major findings of present dissertation are

consistent with its assumptions. For instance, Chapter 2 revealed that action- but not state-oriented

individuals utilize their working memory capacity more efficiently under high compared to low

demands. Moreover, Chapter 3 found that action- compared to state-oriented individuals are better


able to regulate their affective states (cf. Koole & Jostmann, 2004). The regulation of one’s

affective states can be expected to increase the updating of working memory. Finally, Chapter 4

found that improved cognitive control under high demands among action-oriented individuals only

occurred when cognitive control relied heavily on working memory because external prompts were

lacking. To the extent that efficient use of working memory involves updating, the present findings

suggest that action- compared to state-oriented individuals can better update working memory

under high demands.

Theoretical Implications

The present dissertation examined coping with demands using well-established and

validated measures of high-level functioning including the OSPAN task of working memory and

the Stroop task. These tasks have been used in various psychological sub-disciplines such as, for

instance, cognitive psychology or cognitive neuroscience. As such, the present dissertation helps to

bridge the literature on action control and coping with high demands with existing cognitive and

neuropsychological theories of executive functioning. The present dissertation thus indicates that

much is to be gained from a continued theoretical exchange between the areas of intentional action

control, coping with demands, and cognitive neuroscience.

A powerful theoretical approach in cognitive neuroscience is formed by recent neuro-

cognitive models of working memory regulation (Braver & Cohen, 2000). As outlined above, the

process whereby working memory gets updated could provide an integrative theoretical framework

for understanding the effects of action versus state orientation on coping with demanding situations.

Testing the predictions of the updating model in the context of coping with high demands may be

an important task for future research. One important prediction that could be tested during future

research is the notion that coping with high demands depends on how efficient people can update

their working memory. Because updating is linked to the up-regulation of positive affect (or

dopaminergic activity), one can assume an important mediating role of affect regulation during the


coping process (cf. Kuhl, 2000). In the context of the present research, one would predict that affect

regulation mediates the effects of action versus state orientation and high demands on working

memory (cf. Chapter 2) and on cognitive control (cf. Chapter 4). For instance, improved cognitive

control in Stroop tasks among action- compared to state-oriented individuals under high demands

should be mediated by more effective affect regulation among action-oriented individuals.

The mediating role of affect regulation in the effects of action versus state orientation is

theoretically plausible (Braver & Cohen, 2000; Koole & Kuhl, in press; Kuhl, 2000). However, it

may not be easy to detect empirically. A possible problem is that many measures of affect

regulation (e.g., the affective Simon task; De Houwer & Eelen, 1998) interfere with the regulatory

process and therefore decrease the possibility to find effects on subsequent measures of action

control. Likewise, it might be difficult to detect mediating effects of dopaminergic activity because

dopamine cannot be measured directly in human participants.

However, recent research suggests a promising way to assess dopaminergic activity

unobtrusively through its behavioral correlates. Specifically, dopaminergic activity correlates with

spontaneous eye blinks, which makes the assessment of spontaneous eye blinks through video

observation, EMG, or EEG registration a potential way to measure dopaminergic activity (Bacher &

Smotherman, 2004; Dreisbach et al., 2005). Another unobtrusive way to measure dopaminergic

activity may be to assess dopamine gene polymorphisms (i.e., the DRD4 exon III 4/7 genotype;

Dreisbach et al., 2005). Taken together, using unobtrusive measures of dopaminergic activity such

as spontaneous eye blinks or gene polymorphisms may be fruitful way to examine the predicted

affective-dopaminergic mediation effects on coping with high demands.

On a more global level, the insights of the present dissertation may inform different research

fields that examine self-regulation processes that seem closely related to working memory function.

For instance, social psychological research on so-called “ego-depletion” (Muraven & Baumeister,

2000; Schmeichel, Vohs, & Baumeister, 2003), neuro-cognitive research on mental fatigue (Lorist,


Klein, Nieuwenhuis, De Jong, Mulder, & Meijman, 2000; Van der Linden, Frese, & Meijman,

2003), and classic coping research (cf. Hockey, 1997; Zeidner & Endler, 1996) are conceptually

and methodologically distinct areas of research. However, the relevant empirical phenomena

display some important similarities in that they all involve drops in performance under high

demands. One intriguing issue for future research is therefore whether similar or different processes

are involved during coping with, for instance, ego depletion and coping with mental fatigue.

On the basis the present dissertation, it would be informative to consider the role of action

versus state orientation in phenomena such as ego depletion. Typical ego depletion studies measure

how long people can persist on a tiring task, such as squeezing a handgrip (cf. Martijn, Tenbült,

Merckelbach, Dreezens, & De Vries, 2002; Muraven, Tice, & Baumeister, 1998). Such tasks

require participants to inhibit their impulses (e.g., giving up on the physical exercise). However,

they do not necessarily require participants to maintain an explicit intention representation in

memory because the intention is already prompted by the task context (e.g., doing the exercise).

The findings of the present dissertation suggest that state- compared to action-oriented individuals

have no more difficulties to exert cognitive control under high demands when the task intention is

externally prompted (see particularly Chapter 4). As such, the present perspective suggests that

state-oriented individuals may perform equally well or even better than action-oriented individuals

in many ego depletion tasks (cf. Baumann & Kuhl, 2005; Fuhrmann & Kuhl, 1998). Although this

provocative hypothesis remains to be tested empirically, one can conclude that considering the

findings of the present dissertation may contribute intriguing new insights to other research on self-

regulation.

Practical Implications

The findings of the present dissertation also have practical implications. One major set of

findings of the present dissertation is that individuals vary widely in their ability to cope with

demands. The same intervention that may help one group of individuals is likely to undermine the


performance of another group of individuals. Thus, the results of the present dissertation strongly

suggest that individual differences in specific coping abilities should be taken seriously. Imposing a

uniform task structure upon both action- versus state-oriented individuals may well result in sub-

optimal performance for either group of individuals. The most optimal interventions will tailor task

settings to the specific characteristics of the individuals.

State-oriented individuals seem to warrant special attention when it comes to coping with

demands. The present dissertation found repeatedly that state-oriented individuals’ inability to

utilize their mental capacities efficiently under high demands (for instance, see Chapter 2, Studies

2.1 – 2.2). These findings converge with field studies showing that state-oriented individuals’

inability to utilize their mental capacities efficiently under high demands may lead to performance

impairments in many everyday achievement settings such as at work or during one’s study

(Bossong, 1998; Diefendorff, Hall, Lord, & Strean, 2000). The question therefore arises how state-

oriented individuals can be helped to meet their goals more effectively.

Before I turn to more practical considerations, it is important to recognize that high

performance of the individual under high demands is valued especially in modern Western society.

There exist other societies, for instance, who place greater value on virtues such as self-sacrifice or

collective rather than individual performance (cf. Markus & Kitayama, 1991). As such, having a

state orientation may be become more problematic the more a society or culture values high

effectiveness under high demand. The adaptiveness of state orientation may indeed be considered

from an interpersonal or group level. Specifically, it is possible that state-oriented individuals’

hesitation is adaptive on a group level as, for instance, when it prevents the premature

implementation of erroneous group decisions (for initial evidence, see Haschke & Kuhl, 1995;

Olvermann, Metz-Göckel, Hannover, & Pöhlmann, 2004). Practitioners are thus advised to take a

broad perspective on the utility of action versus state orientation, and consider potential tradeoffs


between the interests of the individual and the interests of the broader social system within which

the individual functions.

It is also important to consider whether state orientation can sometimes be adaptive (cf.

Koole, Kuhl, Jostmann, & Vohs, 2005). The present research found that state-oriented individuals

sometimes outperform action-oriented individuals under low demands (for instance, see Chapter 2).

However, it is possible that such performance improvements were not due to greater processing

efficiency (cf. Eysenck & Calvo, 1992; Gray, Burgess, Schaefer, Yarkoni, Larsen, & Braver, 2005)

but to greater effort investment (cf. Kahneman, 1973). The problem with great effort investment is

that it may lead to resource depletion among state-oriented individuals (cf. Hockey, 1997; Muraven

& Baumeister, 2000). As such, the benefits of being state-oriented under low demands may only be

short-lived and backfire when resources are depleted.

In short, it seems likely that many state-oriented individuals will display significant

functional deficits under demanding conditions, deficits that these individuals would gladly do

without. As such, I will consider the potential merits of some strategies that might be used to help

individuals to cope with high demands. One radical strategy may be to eliminate demands

altogether from the lives of state-oriented individuals. This strategy seems difficult to realize,

because demands are an integral aspect of modern life. However, even more moderate strategies

that seek to reduce dermands as much as possible should be applied with caution. Some prior

research research suggests that self-regulatory resources such as action orientation are much like a

skill that has to be learned and trained (Schulte, Hartung, & Wilke, 1997; cf. Muraven, Baumeister,

& Tice, 1999). Thus, although reducing demands is likely to provide temporary relief to state-

oriented individuals, it may at the same time deprive these individuals from the opportunity to

increase their coping skills to deal with demanding situations in other contexts. Consequently, the

strategy of reducing or eliminating demands altogether may ultimately backfire for state-oriented

individuals.


In Chapter 4, Study 4.4, cognitive control among state-oriented individuals was greatly

facilitated by providing them with external prompts. Thus, applied settings may use external

prompts to help state-oriented students or employees to free up their working memory and thus to

maintain their behavioral efficiency under high demands. In more unstructured and unpredictable

situations, however, it will be useful to teach state-oriented individuals strategies to generate their

own idiosyncratic prompts and to form implementation intentions or action plans (cf. Gollwitzer,

1996).

It should further be noted that the functional impairments of state orientation should not

apply for tasks that can be executed on the basis of automatic behavior routines. Some

psychologists have argued that many of people’s daily activities can be performed on automatic and

routine levels (Aarts & Dijksterhuis, 2000; Bargh & Chartrand, 1999). As such, state-oriented

individuals may be able to function well during most daily activities. Moreover, even in cases that

people cannot rely on automatic processing state-oriented individuals may perform similarly well as

action-oriented individuals when external prompts are available (cf. Chapter 4). Because everyday

life is full of external prompts that remind people of what they should do such as traffic signs,

advertisements, calendars, alarm clocks, notebooks etc., state-oriented individuals may experience

little hinder under high demands when such prompts are present.

A final set of questions relates to whether and how state-oriented individuals might be

encouraged to become more action-oriented. According to PSI theory (Kuhl, 2000; cf. Koole, Kuhl,

Jostmann, & Finkenauer, 2006), action versus state orientation is not biologically determined but

develops from childhood experiences with demanding situations. As such, it is theoretically be

possible for a person to change from reacting in a predominantly state-oriented mode in demanding

situations towards a more action-oriented mode of responding (cf. Schulte et al., 1997). One

possible way to make people more action-oriented may be to expose them to moderate levels of

demands in order to inoculate them against higher levels of demands (cf. Bossong, 1998;


Meichenbaum & Deffenbacher, 1988). An effective way to implement such training may be to

focus on specific critical life domains such as work, education, or social relations, rather than trying

to improve someone’s overall action orientation across situations (cf.Diefendorff, Hughes, &

Kamin, 2001).

Concluding Remarks

In modern society, coping effectively with high demands is increasingly necessary for

managing one’s professional, academic, and social lives. The present dissertation found converging

evidence that coping effectiveness is intimately linked to intentional action control. Indeed, action-

oriented individuals were found to cope more effectively with high demands than state-oriented

individuals in terms of the functioning of working memory, affect regulation, and cognitive control.

As such, action versus state orientation represents a key factor in knowing who gets going when the

going gets tough.


Endnotes

1.1. Parts of Chapter 1 and Chapter 5 of the present dissertation are based on Jostmann, N.

B., & Koole, S. L. (2007). Dealing with demands: The role of action versus state orientation. To

appear in R. Hoyle (Ed.), The handbook of personality and self-regulation. Boston, USA:

Blackwell. www.blackwellpublishing.com/

1.2. I use the term “automaticity” to refer to mental and behavioral processes that are

unintended and that elude people’s conscious awareness. The issue of automaticity versus

controllability of human behavior is complex and subject to an ongoing debate, which is beyond the

scope of the present dissertation. The interested reader may further consult Goschke (2003), Kuhl

and Koole (2004), Monsell (1996), Moors and De Houwer (2006), and Wegner and Bargh (1998).

1.3. The terms “cognitive control”, “action control”, “executive control”, “voluntary

control”, “self-control”, “control”, and “volition” have often been used interchangeably in the

literature. In the present dissertation, I use the term “cognitive control” to refer to what Miyake and

colleagues (2000, p. 57) denote as “one’s ability to deliberately inhibit dominant, automatic, or

prepotent responses when necessary”. By contrast, I use the term “action control” more generally to

refer to the mental processes that lead to the formation, maintenance, and implementation of (or

disengagement from) an intention (cf. Kuhl & Goschke, 1994a).

1.4. The term “demand” has elsewhere been used in different ways to refer to conditions in

which people have to perform several mental operations simultaneously or in quick succession

(e.g., Beilock et al., 2004; cf. Pashler, 1994), or to refer to situations that people experience as

potentially threatening for their well-being or self-image (e.g., Blascovich & Mendes, 2000).

1.5. It seems likely that sudden declines in positive affect can have very similar effects on

working memory effectiveness as sustained working memory use. Such declines can either be

triggered directly through, for instance, externally presented aversive stimuli (cf. Study 3.1), or


indirectly through the activation of intention memory during prospective load (cf. Study 4.4; Kuhl,

2000; Kuhl & Kazén, 1999).

1.6. In the present dissertation, the most relevant dimension of the ACS-90 is demand-

related action versus state orientation (AOD). All examples and references are therefore related

specifically to AOD (except for Chapter 3). For the sake of brevity, I use the more general term

action versus state orientation throughout this dissertation to refer to AOD. Discussions of threat-

related action versus state orientation (AOT), another facet of the broader action orientation

construct, can be found in Chapters 2 – 4. The terms “demand-related” and “threat-related” action

versus state orientation were suggested by Koole and Jostmann (2004) to replace the original

“decision-related” and “failure-related” action versus state orientation, respectively (Kuhl, 1994a,

1994b). The new labels fit better with relevant constructs within PSI theory (Kuhl, 2000, 2001).

2.1. Chapter 2 is based on: Jostmann, N. B., & Koole, S. L. (2006). On the waxing and

waning of working memory: Action versus state orientation moderates the impact of demanding

relationships primes on working memory capacity. Personality and Social Psychology Bulletin, 32,

1716-1728. Personality and Social Psychology Bulletin, Vol. 32, (12) © 2006 by Society for

Personality and Social Psychology, Inc. http://psp.sagepub.com/ This article does not exactly

replicate the final version published in the journal Personality and Social Psychology Bulletin. It is

not a copy of the original published article and is not suitable for citation.

2.2. A correlational analysis (Koole, 2005) on undergraduates’ responses (N = 67) on the

ASC-90 (Kuhl, 1994b) and the Regulatory Focus Questionnaire (Higgins et al. 2001) revealed that

AOD correlates moderately with both promotion (r = .42; p < .001) and prevention focus (r = .35; p

< .003).

2.3. A regression approach. revealed a significant (orientation) x (visualization) interaction

on OSPAN in Study 2.1, β = .271, t(73) = 2.35, p < .03, R2 = .08. In Study 2.2, the (orientation) x

(visualization) interaction on the intention superiority effect was also significant, β = .237, t(117) =


2.60, p < .02, R2 = .06. Thus, a regression approach yielded results equivalent to ANOVA approach.

Because a regression approach made it difficult to inspect the absolute means of the dependent

variables, we report the ANOVA results in the main body of Chapters 2 – 4.

2.4. A’ was calculated according to the following formulas: (a) if H > FA, A’ = .5 + (H –

FA) (1 + H – FA)/4H(1 – FA), (b) if H = FA, A’ = .5, and (c) if H < FA, A’ = .5 – (FA – H) (1 + FA

– H)/4FA(1 – H), where H is the hit rate and FA is the false alarm rate.

3.1. Chapter 3 is based on: Jostmann, N. B., Koole, S. L., Van der Wulp, N. Y., &

Fockenberg, D. A. (2005). Subliminal affect regulation: The moderating role of action versus state

orientation. European Psychologist, 10, 209-217. European Psychologist, Vol. 10, (3) © 2005 by

Hogrefe & Huber Publishers http://www.apa.org/journals/epp/ This article does not exactly

replicate the final version published in the journal European Psychologist. It is not a copy of the

original published article and is not suitable for citation.

3.2. A regression approach showed that the critical (AOD) x (prime valence) interaction on

basic affect was significant, β = - .242, t(84) = - 2.25, p < .03.

3.3. We believe that the low reliability coefficient of the basic affect measure can be

explained by a peculiarity of our task. Specifically, task instructions informed participants that the

experimenters were interested in why some combinations of letters were evaluated more positively

than others. Presumably, these instructions had made participants believe that their preferences for

some of the stimuli should be expressed in comparison to other stimuli. As a consequence, a

relatively positive evaluation of a target stimulus might have been followed by a comparatively

negative evaluation of a subsequently presented target stimulus. Given the random order

presentation of the three stimuli, intercorrelations appeared to be rather weak.

4.1. Chapter 4 is based on: Jostmann, N. B., & Koole, S. L. (in press). On the regulation of

cognitive control: Action orientation moderates the impact of situational demands in Stroop

interference tasks. Journal of Experimental Psychology: General. © 2007 by the American


Psychological Association http://www.apa.org/ This article does not exactly replicate the final

version published in the journal European Psychologist. It is not a copy of the original published

article and is not suitable for citation. Note that an earlier version of Chapter 4 included an

additional study, which examined the impact of action versus state orientation and demands on the

gender priming effect (cf. Blair & Banaji, 1996). Because it is not clear whether the gender priming

effect is conceptually equivalent to the Stroop effect (cf. De Houwer, 2003; Kornblum, Hasbroucq,

& Osman, 1990), I dropped the study from the final version of this chapter.

4.2. Study 4.1 also examined the effects of individual differences in achievement motivation

on cognitive control. To this end, we used Elliot and McGregor’s (2001) goal achievement scale,

which differentiates between mastery-related achievement motivation (6 items; Cronbach’s Alpha =

.76), and performance-related achievement motivation (6 items; Cronbach’s Alpha = .68). We

found no significant correlations between action versus state orientation and mastery-related

achievement motivation (r = .14), or performance-related achievement motivation (r = -.10),

respectively. Moreover, a 2 (orientation: action vs. state; between participants) x 2 (mastery

motivation: high vs. low; between participants) ANOVA on Stroop interference in response

latencies revealed a significant main effect of orientation, F(1, 49) = 5.11, p < .03, ηp2 = .09, but no

effects involving mastery motivation, all Fs < 1.7. In addition, a 2 (orientation: action vs. state;

between participants) x 2 (performance motivation: high vs. low; between participants) ANOVA on

Stroop interference in response latencies revealed a significant main effect of orientation, F(1, 49) =

4.53, p < .04, ηp2 = .09, and a marginal effect of performance motivation, F(1, 49) = 3.18, p < .07,

ηp2 = .07. Participants with high performance motivation displayed somewhat more Stroop

interference than participants with low performance motivation (M = 99 ms vs. M = 69 ms). The

interaction term between orientation and performance motivation was not significant, F < 1. In sum,

the effects of action versus state orientation on cognitive control in Study 4.1 were independent of

individual differences in achievement motivation.


4.3. Following Kuhl (1994b), we included measures of threat-related action versus state

orientation (AOT) in Studies 4.1 – 4.4. Across Studies 4.1 – 4.4, we found no interactive effects of

AOT and demands on cognitive control. Therefore, we do not further report on this individual

difference.

4.4. We also examined our data in Studies 4.1 – 4.4 using participants’ responses on the

AOD subscale as a continuous variable in a regression approach. In Study 4.1, the effect of

orientation was significant even when we controlled for OSPAN scores, β = -.28, t(52) = -2.08, p <

.05, R2 = .09. In Study 4.2, we conducted a (trial difficulty) x (task duration) repeated measures

analysis with participants’ responses on the AOD subscale as a covariate. The analysis revealed a

significant interaction between trial difficulty, task duration, and the covariate, F(2, 72) = 4.87, p <

.02, ηp2 = .12. In Study 4.3, the (AOD) x (real-life demand) interaction was significant, β = -.30,

t(76) = -2.64, p < .02, R2 = .11. In Study 4.4, the (AOD) x (prospective load) x (congruency

proportion) interaction was significant, β = .17, t(140) = 2.07, p < .05, R2 = .18. Within the 75%

congruent condition, the (AOD) x (prospective load) interaction also reached the level of

significance, β = -.253, t(68) = -2.10, p < .05, R2 = .07.

4.5. In Study 4.4, we deviated from the normative split because of an otherwise

disproportionally unequal distribution of participants across cells. We based the classification of

participants as either action- versus state-oriented on the empirical rather than the normative

median. Participants with seven or less action-oriented responses on the AOD subscale were

classified as state-oriented, whereas the remaining participants were classified as action-oriented.

Rerunning the main analyses after exclusion of the mid-group of participants (i.e. participants who

had given six or seven action-oriented responses; n = 29) left the results intact. Specifically, an

(orientation) x (prospective load) x (congruency proportion) ANOVA revealed the predicted three-

way interaction, F(1, 105) = 4.43, p < .04, ηp2 = .04. A separate (orientation) x (prospective load)

ANOVA within the 75% congruent condition yielded the predicted two-way interaction, F(1, 49) =


3.90, p = .054, ηp2 = .07. The results in Study 4.4 were thus not dependent on the use of the

empirical versus the normative median. Moreover, when we conducted regression analyses on the

entire sample, we obtained equivalent results to an ANOVA approach (see Footnote 4.4).

5.1. Note that working memory should not erase old information that is still relevant.

Moreover, working memory should also not allow new but irrelevant information to enter. These

two situations are phenomenologically experienced as forgetfulness and distraction, respectively.

Working memory and the dopaminergic system thus have the paradoxical task of safeguarding

against irrelevant information but still remaining open for relevant information. How this task is

solved has not yet been fully understood (cf. Goschke, 2003). One suggestive possibility is the

existence of two antagonistic dopamine systems that become activated when information is

categorized as irrelevant or relevant, respectively. In line with this idea, Cohen and colleagues

(2002) recently suggested that so-called D1-receptors regulate tonic increases in dopamine thereby

protecting information in working memory. By contrast, D2-receptors regulate phasic increases in

dopamine thereby facilitating the updating of working memory with new information (cf.

Durstewitz et al., 1999).


Summary

High demands are a pervasive condition in many people’s professional, academic, and social

lives. High demands occur, for instance, when people have many pressing affairs simultaneously on

their mind. Because high demands can make people less successful at reaching their goals, it is

important to understand how people can cope effectively with high demands. The present

dissertation examines coping with high demands from an action control perspective (Kuhl, 1984,

2000). Action control refers to the mental processes that are involved in the formation,

maintenance, and implementation of (or disengagement from) an intention. Such processes rely on a

high-level working memory system that is partly located in the prefrontal cortex.

In the present dissertation, high demands refer to conditions that are characterized by

sustained use of working memory. Sustained use of working memory can lead to a transient decline

of its effectiveness. Because working memory effectiveness is a key function of action control,

strong action control skills may be an important factor in coping with high demands. Accordingly,

the present dissertation tested the hypothesis that individuals who are highly skilled at action

control, or action-oriented individuals, cope more effectively with high demands than individuals

who are less skilled at action control, or state-oriented individuals.

In three series of studies, I experimentally induced high versus low levels of demands or I

measured participants’ level of demand in their real lives. Coping effectiveness among action-

versus state-oriented individuals was measured across three major psychological domains including

cognition, affect, and behavior. In a first line of research (Chapter 2), action- compared to state-

oriented individuals were found to be better able at cognitive shielding as reflected by the efficiency

with which people can make use of their working memory capacity under high demands.

Specifically, action-oriented individuals displayed more efficient use of their working memory

capacity in an “operation span task” (Study 2.1) and a “postponed intention task” (Study 2.2) under

high compared to low demands. Among state-oriented individuals, I found the reversed pattern such


that high demands led to less efficient use of working memory than low demands. A second line of

research (Chapter 3) revealed that action-oriented individuals are also better able at affective

shielding as reflected by their positive affective reactions in response to subliminally primed

negative affect in comparison with state-oriented individuals’ reactions (Study 3.1).

In a third and final line of research, action- compared to state-oriented individuals were also

found to be better able at behavioral shielding (Chapter 4) as reflected by the ability to implement a

difficult intention. Specifically, in a series of “Stroop color evaluation tasks”, action-oriented

individuals were found to be better able to override a strong but inappropriate response tendency

(Studies 4.1 – 4.4). Across all three lines of research, improved performance among action-

compared to state-oriented individuals was found only under high but not under low demanding

conditions indicating that action compared to state orientation reflects better coping abilities rather

than better overall mental skills across conditions.

To integrate the findings, the present dissertation proposes the “updating model” of action

control and coping with high demands (Chapter 5). Based on recent neuro-cognitive theories of

working memory regulation (Braver & Cohen, 2000; Kuhl, 2000), the updating model suggests that

high demands inhibit the neuro-cognitive pathway between working memory and behavioral output

systems thereby rendering an updating of information between these two systems difficult. When

updating fails, working memory utilization becomes less efficient, and intentions become less

available to guide behavioral output. Furthermore, pathway inhibition is linked to decreased

dopaminergic activity and reduced positive affect. Taken together, failure to update working

memory is la likely reason for detrimental effects of high demands on different psychological

domains including cognition, affect, and behavior.

The updating model further suggests that action compared to state orientation facilitates the

updating function of working memory. A likely way how action orientation restores the updating

function is by upregulating positive affect (cf. Koole & Jostmann, 2004). The empirical test of this


assumption provides an important task for future research on effective coping with high demands.

Chapter 5 provides some ideas about how this task may be accomplished. In sum, the updating

model integrates the findings of the present dissertation with contemporary theories of working

memory regulation, and provides some promising perspectives for further research.


Samenvatting

Veel mensen komen tijdens hun werk, in hun studie, en in hun sociale leven geregeld onder

hoge psychische druk te staan. Hoge psychische druk ontstaat bijvoorbeeld als mensen veel

dringende aangelegenheden tegelijkertijd moeten regelen. Als gevolg hiervan zijn mensen vaak

minder succesvol in het bereiken van hun doelen. Gezien deze en andere potentiële gevolgen is het

belangrijk om te begrijpen hoe mensen op een effectieve manier kunnen omgaan met belastende

situaties die hoge psychische druk veroorzaken. Dit proefschrift onderzoekt deze vraag vanuit het

perspectief van de actie-controle theorie (Kuhl, 1984, 2000). Actie-controle verwijst naar de

mentale processen die betrokken zijn bij het vormen, vasthouden (of loslaten), en uitvoeren van

gedragsintenties. Deze processen maken allemaal gebruik van het zogenaamde werkgeheugen, een

geavanceerd mentaal systeem dat ten dele gelokaliseerd is in de prefrontale hersengebieden.

Een belangrijk kenmerk van belastende situaties is dat mensen aanhoudend gebruik moeten

maken van hun werkgeheugen. Door aanhoudend gebruik kan het werkgeheugen echter tijdelijk

uitgeput raken. Omdat het werkgeheugen een belangrijke functie heeft voor actie-controle, is het

aannemelijk dat mensen beter met druk kunnen omgaan naarmate zij bekwamer zijn in actie-

controle. In dit proefschrift heb ik de hypothese onderzocht dat mensen die bekwaam zijn in actie-

controle, of actiegeoriënteerde mensen, effectiever kunnen omgaan met psychische druk dan

mensen die minder bekwaam zijn in actie-controle, of toestandsgeoriënteerde mensen.

In drie reeksen onderzoeken heb ik bij actie- en toestandsgeoriënteerde proefpersonen hoge

of lage druk experimenteel geïnduceerd of de mate van druk in hun persoonlijke levens gemeten.

De effectiviteit in het omgaan met druk heb ik vervolgens op drie niveaus van psychologisch

functioneren gemeten: cognitie, affect, en gedrag. In een eerste reeks onderzoeken naar zogenaamde

cognitieve bescherming (Hoofdstuk 2) heb ik gevonden dat de efficiëntie van het werkgeheugen bij

actiegeoriënteerden beter is onder hoge dan onder lage druk, zoals bleek uit hun prestatie in de

zogenaamde “operatie-spanne taak” (Studie 2.1) en in de “uitgestelde intentie taak” (Studie 2.2). Bij


toestandsgeoriënteerden vond ik daarentegen efficiënter gebruik van het werkgeheugen onder lage

dan onder hoge druk. In een tweede onderzoekslijn naar affectieve bescherming (Hoofdstuk 3) vond

ik dat actiegeoriënteerden positievere affectieve reacties gaven dan toestandsgeoriënteerden nadat

ik bij hen onbewust negatief affect had geïnduceerd (Studie 3.1).

In een derde en laatste onderzoekslijn naar gedragsmatige bescherming (Hoofdstuk 4)

bleken actiegeoriënteerden moeilijke intenties beter uit te kunnen voeren dan

toestandsgeoriënteerden. In een reeks zogenaamde “Stroop kleurbenoemingstaken” (Studies 4.1 –

4.4) konden actiegeoriënteerden onder hoge druk een dominante maar onbedoelde respons beter

onderdrukken en een bedoelde respons beter uitvoeren. In alle drie de onderzoekslijnen bleken

actiegeoriënteerden alleen beter te presteren dan toestandsgeoriënteerden wanneer de druk hoog

was. Deze observatie suggereert dat actiegeoriënteerden niet onder alle omstandigheden over

grotere mentale capaciteiten beschikken dan toestandsgeoriënteerden.

Om de bevindingen van dit proefschrift theoretisch te integreren formuleer ik in Hoofdstuk 5

het “actualiseringmodel” van actie-controle en omgaan met psychische druk. Het model bouwt

voort op recente neuro-cognitieve theorieën over de regulatie van het werkgeheugen (Braver &

Cohen, 2000; Kuhl, 2000). Eén belangrijke veronderstelling van het actualiseringmodel is dat hoge

druk de neuro-cognitieve verbinding tussen het werkgeheugen en gedragsuitvoerende systemen

verzwakt. Door deze verzwakking wordt de actualisering van informatie in het werkgeheugen

moeilijker zodat irrelevante informatie niet uit het werkgeheugen verwijderd kan worden en

relevante informatie niet voldoende verwerkt kan worden. Als gevolg hiervan nemen de efficiëntie

van het werkgeheugen en de beschikbaarheid van intenties voor de gedragscontrole af. Daarnaast

wordt de activiteit van de neuronale overbrenger dopamine, en daardoor de opwekking van positief

affect, verminderd. Met andere woorden, als het werkgeheugen onder hoge druk niet meer

voldoende geactualiseerd kan worden, neemt het psychologische functioneren op cognitief,

affectief, en gedragsniveau af.


Een tweede veronderstelling van het actualiseringmodel is dat actieoriëntatie de

actualisering van het werkgeheugen bevordert. Dit gebeurt waarschijnlijk door de opwekking van

positief affect tijdens hoge psychische druk (zie Koole & Jostmann, 2004). De empirische test van

deze aanname vormt een belangrijke taak voor toekomstig onderzoek naar de gevolgen van hoge

druk. In Hoofdstuk 5 geef ik een aantal suggesties hoe dit gedaan zou kunnen worden. Samengevat

integreert het actualiseringmodel de bevindingen van dit proefschrift met hedendaagse theorieën

over de regulatie van het werkgeheugen, en biedt het een aantal veelbelovende perspectieven voor

toekomstig onderzoek naar psychische druk.


Zusammenfassung

Viele Menschen sind im Beruf, in der Ausbildung und in ihrem sozialen Umfeld hohen

mentalen Belastungen ausgesetzt, etwa dann, wenn viele dringende Angelegenheiten zur gleichen

Zeit erledigt werden müssen. Weil es bei hoher Belastung oft schwierig ist, persönliche Ziele zu

erreichen, stellt sich die Frage, wie Menschen auf eine effektive Weise mit Belastungen umgehen

können. In der vorliegenden Dissertation wird diese Frage aus der Perspektive der

Handlungskontrolltheorie (Kuhl, 1984, 2000) untersucht. Die Handlungskontrolltheorie beschreibt

die mentalen Prozesse, die beim Formen, Festhalten (oder Loslassen) und Ausführen von

Verhaltensabsichten beteiligt sind. Kern dieser Prozesse ist das so genannte Arbeitsgedächtnis, ein

komplexes mentales System, das zum Teil in präfrontalen Gehirnarealen lokalisiert ist.

Während hoher Belastungen wird das Arbeitsgedächtnis fortlaufend in Anspruch

genommen. Bei längerer Inanspruchnahme lässt seine Effektivität oft nach. Da das

Arbeitsgedächtnis eine wichtige Funktion innerhalb der Handlungskontrolle erfüllt, ist zu vermuten,

dass der effektive Umgang mit Belastungen eine starke Handlungskontrolle voraussetzt. In der

vorliegenden Dissertation habe ich daher die Hypothese getestet, dass Menschen, die über eine

starke Handlungskontrolle verfügen (Handlungsorientierte), effektiver mit Belastungen umgehen

können als Menschen, deren Handlungskontrolle schwächer ist (Lageorientierte).

In drei Untersuchungenreihen habe ich hohe oder niedrige Belastungen bei handlungs- und

lageorientierten Probanten experimentell induziert beziehungsweise das Belastungsniveau in deren

Alltagsleben erfasst. Die Effektivität im Umgang mit hohen Belastungen habe ich auf drei

verschiedenen psychologischen Funktionsebenen untersucht: Denken, Fühlen, und Handeln. In

einer ersten Reihe von Untersuchungen zur Abschirmung des Denkens (Kapitel 2) zeigte sich an

Hand der so genannten “Operationsspannenaufgabe” (Studie 2.1) und der “Aufgeschobenen

Absichtssaufgabe” (Studie 2.2), dass die Effizienz des Arbeitsgedächtnis bei Handlungsorientierten

unter hohen Belastungen größer ist als unter niedrigen Belastungen. Bei Lageorientierten


beobachtete ich den umgekehrten Effekt, nämlich dass die Effizienz des Arbeitsgedächtnis unter

hohen Belastungen kleiner ist als unter niedrigen Belastungen. In einer zweiten Untersuchungseihe

zur Abschirmung des Fühlens (Kapitel 3) beobachtete ich bei Handlungsorientierten positivere

affektive Reaktionen als bei Lageorientierten, nachdem ich zuvor unterschwellig einen negativen

Affekt induziert hatte (Studie 3.1).

In einer dritten und letzten Reihe von Untersuchungen zur Abschirmung des Handelns

(Kapitel 4) zeigte sich, dass Handlungsorientierte schwierige Verhaltensabsichten bei hoher

Belastung besser ausführen können als Lageorientierte. In einer Reihe so genannter “Stroop-

Farbbenennungsaufgaben” (Studien 4.1 – 4.4) konnten Handlungsorientierte eine dominante, aber

unbeabsichtigte Verhaltensneigung besser unterdrücken, um statttdessen eine beabsichtigte

Handlung auszuführen. In allen drei Untersuchungsreihen war die Leistung Handlungsorientierter

nur bei hoher Belastung besser als die der Lageorientierten. Dieses zeigt, dass Handlungsorientierte

nicht in jeder Situation über größere mentale Fähigkeiten verfügen als Lageorientierte.

Um die Resultate der vorliegenden Dissertation theoretisch zu integrieren, formuliere ich in

Kapitel 5 das “Aktualisierungsmodell” von Handlungskontrolle im Umgang mit hoher Belastung.

Das Modell basiert auf aktuellen neuro-kognitiven Theorien des Arbeitsgedächtnis (Braver &

Cohen, 2000; Kuhl, 2000). Eine Grundannahme des Aktualiserungsmodells ist, dass hohe

Belastungen die neuro-kognitive Verbindung zwischen dem Arbeitsgedächtnis und

verhaltensausführenden Systemen vorübergehend abschwächt. Durch diese Schwächung kann das

Arbeitsgedächtnis nicht mehr ausreichend aktualisert werden, so dass irrelevante Information

aktiviert bleibt, während relevante Information nicht ausreichend verarbeitet werden kann. Als

Folge hiervon nehmen die Effizienz des Arbeitsgedächtnis und damit der Einfluss von Absichten

auf die Verhaltenssteuerung ab. Eine weitere Folge ist, dass sich die Aktivität des neuronalen

Botenstoffs Dopamin und damit die Generierung positiven Affekts verringert. Mit anderen Worten,


wenn das Arbeitsgedächtnis unter hohen Belastungen nicht mehr ausreichend aktualisiert werden

kann, hat dies die Beeinträchtigung des Denkens, Fühlens, und Handelns zur Folge.

Eine weitere Grundannahme des Aktualisierungsmodells ist, dass Handlungsorientierung die

Aktualisierung des Arbeitsgedächtnis erleichtert. Dies geschieht wahrscheinlich durch die

Generierung positiven Affekts bei Handlungsorientierten als Reaktion auf hohe Belastungen (siehe

Koole & Jostmann, 2004). Der empirische Test dieser Annahme ist eine wichtige Aufgabe für die

weitere Erforschung der Folgen hoher Belastungen. In Kapitel 5 unterbreite ich eine Reihe von

Vorschlägen, wie dieser Test erbracht werden kann. Zusammengefasst lässt sich sagen, dass das

Aktualisierungsmodell es ermöglicht, die vorliegenden Ergebnisse mit aktuellen Theorien des

Arbeitsgedächtnisses zu integrieren, und sich daraus vielversprechende Perspektiven für die

zukünftige Erforschung mentaler Belastungen ergeben.


References

Aarts, H., & Dijksterhuis, A. (2000). Habits as knowlegde structures: Automaticity in goal-directed

behavior. Journal of Personality and Social Psychology, 78, 53-63.

Ach, N. (1910). Über den Willensakt und das Temperament [On the act of willing and the

temperament]. Leipzig, Germany: Quelle & Meyer.

Ajzen, I. (1988). Attitudes, personality, and behavior. Chicago, IL: Dorsey.

Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press.

Ashby, F. G., Isen, A. M., & Turken, A. U. (1999). A neuropsychological theory of positive affect

and its influence on cognition. Psychological Review, 106, 529-550.

Aronoff, J., Barclay, A. M., & Stevenson, L. A. (1988). The recognition of threatening facial

stimuli. Journal of Personality and Social Psychology, 54, 647-655.

Ashcraft, M. H. & Kirk, E. P. (2001). The relationships among working memory, math anxiety, and

performance. Journal of Experimental Psychology: General, 130, 224-237.

Bacher, L. F., & Smotherman, W. P. (2004). Spontaneous eye blinking in human infants: A review.

Wiley Interscience, 95-102.

Baddeley, A. (1986). Working memory. Oxford, UK: Oxford University Press.

Baddeley, A. D., & Hitch, G. J. (1974). Working memory. In G. B. Bower (Ed.), The psychology of

learning and motivation: Advances in research and theory (Vol. 8, pp. 47 – 89). New York:

Academic Press.

Baldwin, M. W., & Sinclair, L. (1996). Self-esteem and “if…then” contingencies of interpersonal

acceptance. Journal of Personality and Social Psychology, 71, 1130-1141.

Baldwin, M. W., Carrell, S. E., & Lopez, D. F. (1990). Priming relationship schemas: My advisor

and the Pope are watching me from the back of my mind. Journal of Experimental Social

Psychology, 26, 435-454.


Bandura, A. (1991). Social cognitive theory of self-regulation. Organizational Behavior and

Human Decision Processes, 50, 248-287.

Bargh, J. A. (1994). The four horsemen of automaticity: Awareness, intention, efficiency, and

control in social cognition. In R. S. Wyer & T. K. Srull (Eds.), Handbook of social

cognition, (2nd Ed., Vol. 1 Basic Processes, pp. 1-40).

Bargh, J. A., & Chartrand, T. L. (1999). The unbearable automaticity of being. American

Psychologist, 54, 462-479.

Baumann, N., & Kuhl, J. (2002). Intuition, affect, and personality: Unconscious coherence

judgments and self-regulation of negative affect. Journal of Personality and Social

Psychology, 83, 1213-1223.

Baumann, N., & Kuhl, J. (2005). How to resist temptation: The effects of external control versus

autonomy support. Journal of Personality, 73, 443-470.

Baumann, N., Kaschel, R., & Kuhl, J. (2005). Affect regulation and motive-incongruent

achievement goals: Antecedents of subjective well-being and symptom formation. Journal

of Personality and Social Psychology, 89, 781-799.

Baumeister, R. F. (1984). Choking under pressure: Self-consciousness and paradoxical effects of

incentives on skillful performance. Journal of Personality and Social Psychology, 46, 610-

620.

Baumeister, R. F., Heatherton, T. F., & Tice, D. M. (1994). Losing control: How and why people

fail at self-regulation. San Diego, CA: Academic.

Baumeister, R. F., & Vohs, K. D. (2004). Handbook of self-regulation: Research, theory, and

applications. New York, NY: Guilford.

Bayer, U. C., Ferguson, M. J., & Gollwitzer, P. M. (2003). Voluntary action from the perspective of

social-personality psychology. In S. Maasen, W. Prinz, & G. Roth (Eds.), Voluntary action:

brain, minds, and sociality (pp.86-107). Oxford, UK: Oxford University Press.


Beilock, S. L., Kulp, C. A., Holt, L. E., & Carr, T. H. (2004). More on the fragility of performance:

Choking under pressure in mathematical problem solving. Journal of Experimental

Psychology: General, 133, 584-600.

Berridge, K. C. & Winkielman, P. (2003). What is an unconscious emotion? (The case for

unconscious “liking”). Cognition and Emotion, 17, 181-211.

Beswick, G., & Mann. L. (1994). State orientation and procrastination. In J. Kuhl & J. Beckmann

(Eds.), Volition and personality, (pp. 391-396). Göttingen, Germany: Hogrefe & Huber.

Blair, I. V., & Banaji, M. R. (1996). Automatic and controlled processes in stereotype priming.

Journal of Personality and Social Psychology, 70, 1142-1163.

Blascovich, J., & Mendes, B. (2000). Challenge and threat appraisals: The role of affective cues. In

J. P. Forgas (Ed.). Feeling and thinking: The role of affect in social cognition (pp. 59-82).

New York, NY: Cambridge University Press.

Blunt, A., & Pychyl, T. A. (1998). Volitional action and inaction in the lives of undergraduate

students: State orientation, boredom, and procrastination. Personality and Individual

Differences, 248, 837-846.

Boksem, M. A. S. (2006). Mental fatigue: Costs and benefits. Doctoral dissertation, University of

Groningen, the Netherlands.

Boksem, M. A. S., Lorist, M. M., & Meijman, T. F. (2005). Effects of mental fatigue on attention:

An ERP study. Cognitive Brain Research, 25, 106-117.

Bossong, B. (1998). Stress und Handlungskontrolle [Stress and action control]. Göttingen,

Germany: Hogrefe.

Botvinick, M. M., Braver, T. S., Barch, D. M., Carter, C. S., & Cohen, J. D. (2001). Conflict

monitoring and cognitive control. Psychological Review, 108, 624-652.

Braver, T. S. & Cohen, J. D. (2000). On the control of control: The role of dopamine in regulating

prefrontal function and working memory. In S. Monsell & J. Driver (Eds.), Attention and


performance XVIII: Control of cognitive processes (713-737). Cambridge, MA: MIT Press.

Braver, T. S., Barch, D. M., Keys, B. A., Carter, C. S., Cohen, J. D., Kaye, J. A., Janowsky, J. S.,

Taylor, S. F., Yessavage, J. A., Mumenthaler, M. S., Jagust, W. J., & Reed, B. R. (2001).

Context processing in older adults: Evidence for a theory relating cognitive control to

neurobiology in healthy aging. Journal of Experimental Psychology: General, 130, 746-763.

Brehm, J. W., & Self, E. A. (1989). The intensity of motivation. Annual Review of Psychology, 40,

109-131.

Broadbent, D. E., Cooper, P. F., FitzGerald, P., & Parkes, K. R. (1982). The cognitive failures

questionnaire (CFQ) and its correlates. British Journal of Clinical Psychology, 21, 1-16.

Brunstein, J. C. (2001). Persönliche Ziele und Handlungsorientierung: Wer bindet sich an

realistische und bedürfniskongruente Ziele? [Personal goals and action orientation: Who

commits to realistic and need-congruent goals?] Zeitschrift für Differentielle und

Diagnostische Psychologie, 22, 1-12.

Brunstein, J. C., & Olbrich, E. (1985). Personal helplessness and action control: Analysis of

achievement-related cognitions, self-assessments, and performance. Journal of Personality

and Social Psychology, 48, 1540-1551.

Cacioppo, J. T., Priester, J. R., & Bernston, G. G. (1993). Rudimentary determinants of attitudes.

Arm flexion and extension have differential effects on attitudes. Journal of Personality and

Social Psychology, 65, 5-17.

Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive and negative affect: A

control-process view. Psychological Review, 97, 19-35.

Carver, C. S., & Scheier, M. F. (2005). Engagement, Disengagement, Coping, & Catastrophe. In A.

J. Elliot, & C. S. Dweck (Eds.), Handbook of competence and motivation (pp. 527-547).

New York, NY: Guilford.


Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and the affective

responses to impending reward and punishment: The BIS/BAS scales. Journal of

Personality and Social Psychology, 67, 319-333.

Chartrand, T. L., & Bargh, J. A. (1996). Automatic activation of impression formation and

memorization goals: Nonconscious goal priming reproduces effects of explicit task

instructions. Journal of Personality and Social Psychology, 71, 464-478.

Chartrand, T. L., Van Baaren, R. B., & Bargh, J. A. (2006). Linking automatic evaluation to mood

and information processing style: Consequences for experienced affect, impression

formation, and stereotyping. Journal of Experimental Psychology: General, 135, 70-77.

Chen, M., & Bargh, J. A. (1999). Consequences of automatic evaluation: Immediate behavioral

predispositions to approach or avoid the stimulus. Personality and Social Psychology

Bulletin, 25, 215-224.

Chen, S., Boucher, H. C., & Tapias, M. P. (2006). The relational self revealed: Integrative

conceptualization and implications for interpersonal life. Psychological Bulletin, 132, 151-

179.

Cohen, J. D., Barch, D. M., Carter, C., & Servan-Schreiber, D. (1999). Context-processing deficits

in schizophrenia: Converging evidence from three theoretically motivated cognitive tasks.

Journal of Abnormal Psychology, 108, 120-133.

Cohen, J. D., Braver, T. S., & Brown, J. W. (2002). Computational perspectives on dopamine

function in prefrontal cortex. Current Opinion in Neurobiology, 12, 223-229.

Cohen, J. D., Braver, T. S., & O'Reilly, R. C. (1998) A computational approach to prefrontal cortex,

cognitive control, and schizophrenia: Recent developments and current challenges. In A. C.

Roberts, T. W. Robbins, & L. Weiskrantz (Eds.), The prefrontal cortex: Executive and

cognitive functions (pp. 195-220). New York, NY: Oxford.


Cohen, J. D., Dunbar, K., & McClelland, J. L. (1990). On the control of automatic processes: A

parallel distributed processing account of the Stroop effect. Psychological Review, 97, 322-

361.

Conway, A. R . A., Cowan, N., & Bunting, M. F. (2001). The cocktail party phenomenon revisited:

The importance of working memory capacity. Psychonomic Bulletin and Review, 8, 331-

335.

Daneman, M., & Carpenter, P. A. (1980). Individual differences in working memory and reading.

Journal of Verbal Learning and Verbal Behavior, 19, 450-466.

De Dreu, C. K. W. (2003). Time pressure and closing of the mind in negotiation. Organizational

Behavior and Human Decision Processes, 91, 280-295.

De Houwer, J. (2003). A structural analysis of indirect measures of attitudes. In J. Musch & K. C.

Klauer (Eds.), The psychology of evaluation: Affective processes in cognition and emotion,

(pp. 219-244). London, UK: Erlbaum.

De Houwer, J., & Eelen, P. (1998). An affective variant of the Simon paradigm. Cognition and

Emotion, 12, 45-85.

De Jong, R., Berendson, E., & Cools, R. (1999). Goal neglect and inhibitory limitations:

Dissociable causes of interference effects in conflict situations. Acta Psychologica, 101,

379-394.

Dibbelt, S. (1997). Wechseln und Beibehalten von Zielen als Subfunktionen der Handlungskontrolle

[Change and maintenance of goals as functional components of action control]. Doctoral

dissertation, University of Osnabrück, Germany.

Diefendorff, J. M. (2004). Examination of the roles of action-state orientation and goal orientation

in the goal-setting and performance process. Human Performance, 17, 375-395.

Diefendorff, J. M. (2006). Correlations between the ACS-90 and the Big Five personality variables.

Unpublished dataset, University of Colorado at Denver, CO.


Diefendorff, J. M., Hall, R. J., Lord, R. G., & Strean, M. L. (2000). Action-state orientation:

Construct validity of a revised measure and its relationship to work-related variables.

Journal of Applied Psychology, 85, 250-263.

Diefendorff, J. M., Hughes, R. R., & Kamin, A. M. (2001). Development of an action-state

orientation measure for the workplace. Unpublished Manuscript, University of Colorado at

Denver, CO.

Diefendorff, J. M., Lord, R. G., Hepburn, E. T., Quickle, J. S., Hall, R. J., & Sanders, R. E. (1998).

Perceived self-regulation and individual differences in selective attention. Journal of

Experimental Psychology: Applied, 4, 228-247.

Diefendorff, J. M., Richard, E. M., & Gosserand, R. H. (2006). Examination of situational and

attitudinal moderators of the hesitation and performance relation. Personnel Psychology, 59,

365-393.

Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & Van Baaren, R. B. (2006). On Making the Right

Choice: The Deliberation-Without-Attention Effect. Science, 311, 1005-1007.

Dreisbach, G., & Goschke, T. (2004). How positive affect modulates cognitive control: Reduced

perseveration at the cost of increased distractibility. Journal of Experimental Psychology:

Learning, Memory, & Cognition, 30, 343-353.

Dreisbach, G., Müller, J., Goschke, T., Strobel, A., Schulze, K., Lesch, KP., & Brocke, B. (2005).

Dopamine and cognitive control: The influence of spontaneous eyeblink rate and dopamine

gene polymorphisms on perseveration and distractibility. Behavioral Neuroscience, 2, 483-

490.

Duckworth, K. L., Bargh, J. A., Garcia, M. T., & Chaiken, S. (2002). The automatic evaluation of

novel stimuli. Psychological Science, 13, 513-519.

Duncan, J. (1995). Attention, intelligence, and the frontal lobes. In M. S. Gazzaniga (Ed.), The

cognitive neurosciences (pp. 721-733). Cambridge, MA: MIT.


Duncan, J., Emslie, H., Williams, P., Johnson, R., & Freer, C. (1996), Intelligence and the frontal

lobe: The organization of goal-directed behavior. Cognitive Psychology, 30, 257-303.

Durstewitz, D., Kelc, M., & Güntürkün, O. (1999). A neurocomputational theory of the

dopaminergic modulation of working memory functions. The Journal of Neuroscience, 19,

2807-2822.

Dweck, C. S. (1986). Motivational processes affecting learning. American Psychologist, 41, 1040-

1048.

Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Orlando, FL: Harcourt Brace

Jovanovich.

Elliot, A. J., & McGregor, H. A. (2001). A 2 X 2 achievement goal framework. Journal of


Engle, R. W., Kane, M. J., & Tuholski, S. W. (1999). Individual differences in working memory

capacity and what they tell us about controlled attention, general fluid intelligence, and

functions of the prefrontal cortex. In A. Miyake, & P. Shah (Eds.), Models of working

memory: Mechanisms of active maintenance and executive control (pp. 102-134). New

York, NY: Cambridge University Press.

Engle, R. W., Tuholski, S. W., Laughlin, J. E., & Conway, A. R. A. (1999). Working memory,

short-term memory, and general fluid intelligence: A latent-variable approach. Journal of

Experimental Psychology: General, 128, 309-331.

Eysenck, M. W., & Calvo, M. G. (1992) Anxiety and performance: The processing efficiency

theory. Cognition and Emotion, 6, 409-434.

Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986). On the automatic

activation of attitudes. Journal of Personality and Social Psychology, 50, 229-238.

Feldman Barrett, L., Tugade, M. M., & Engle, R. W. (2004). Individual differences in working

memory capacity and dual-process theories of the mind. Psychological Bulletin, 130, 553-


573.

Ferguson, M. J., & Bargh, J. A. (2004). How social perception can automatically influence

behavior. Trends in Cognitive Science, 8, 33-39.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, & behavior: An introduction to theory

and research. Reading, MA: Addison-Wesley.

Fitzsimons, G. M., & Bargh, J. A. (2003). Thinking of you: Nonconscious goals associated with

relationships partners. Journal of Personality and Social Psychology, 84, 148-164.

Friedman, N. P., & Miyake, A. (2004). The reading span test and its predictive power for reading

comprehension ability. Journal of Memory and Language, 51, 136-158.

Fuster, J. M. (1995). Memory and planning: Two temporal perspectives of frontal lobe function. In

H. H. Jasper, S. Riggio, & P. S. Goldman-Rakic (eds.), Epilepsy and the functional anatomy

of the frontal lobe (pp. 9-18). New York, NY: Raven.

Gathercole, S. E., & Pickering, S. J. (2000). Working memory deficits in children with low

achievements in the national curriculum at 7 years of age. British Journal ofEducational


Goldberg, L. R. (1992). The development of markers for the Big-Five factor structure.

Psychological Assessment, 4, 26-42.

Gollwitzer, P. M. (1993). Goal Achievement: The role of intentions. European Review of Social


Gollwitzer, P. M. (1996). The volitional benefits of planning. In P. M. Gollwitzer & J. A. Bargh

(Eds.), The psychology of action (pp. 287-312). New York: Guilford.

Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American


Goschke, T. (2003). Voluntary action and cognitive control from a cognitive neuroscience

perspective. In S. Maasen, W. Prinz, & G. Roth (Eds.), Voluntary action: brain, minds, and


sociality (pp.49-85). Oxford: Oxford University Press.

Goschke, T., & Kuhl, J. (1993). Representation of intentions: Persisting activation in memory.

Journal of Experimental Psychology: Learning, Memory, and Cognition, 19, 1211-1226.

Goschke, T., & Kuhl, J. (1996). Remembering what to do: Explicit and implicit memory for

intentions. In M. Brandimonte, G. O. Einstein, & M. McDaniel (Eds.), Prospective memory:

Theory and applications (pp. 53-92). Mahwah, NJ. Erlbaum.

Gray, J. R. (2001). Emotional modulation of cognitive control: Approach-withdrawal states double-

dissociate spatial from verbal two-back task performance. Journal of Experimental


Gray, J. R., Burgess, G. C., Schaefer, A., Yarkoni, T., Larsen, R. J., Braver, T. S. (2005). Affective

personality differences in neural processing efficiency confirmed using fMRI. Cognitive,

Affective, & Behavioral Neuroscience, 5, 182-190.

Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes:

Implications for affect, relationships, and well-being. Journal of Personality and Social


Haschke, R., & Kuhl, J. (1995). Frust und Fliegen [Frustration and flying]. Aeromed Info, 6, 1-2.

Heckhausen, H., & Strang, H. (1988). Efficiency under record performance demands: Exertion

control – An individual difference variable? Journal of Personality and Social Psychology,

55, 489-498.

Hertel, P. T., & Hardin, T. S. (1990). Remembering with and without awareness in a depressed

mood: Evidence of deficits in initiative. Journal of Experimental Psychology: General, 119,

45-59.

Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Review,

94, 319-340.

Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational principle. In


M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 30, pp. 1-46). San

Diego: Academic Press.

Higgins, E. T., Friedman, R. S., Harlow, R. E., Idson, L. C., Ayduk, O. N., Taylor, A. (2001).

Achievement orientations from subjective histories of success: Promotion pride versus

prevention pride. European Journal of Social Psychology, 31, 3-23.

Hockey, G. R. J. (1997). Compensatory control in the regulation of human performance under

stress and high workload: A cognitive-energetical framework. Biological Psychology, 45,

73-93.

James, W. (1890). The principles of psychology. London, UK: Macmillan.

Jostmann, N. B. (2006). Correlations between the ACS-90 and related personality variables.

Unpublished dataset, Vrije Universiteit Amsterdam, the Netherlands.

Jostmann, N. B., & Koole, S. L. (2006). On the waxing and waning of working memory: Action

versus state orientation moderates the impact of demanding relationship primes on working

memory capacity. Personality and Social Psychology Bulletin, 32, 1716-1728.

Jostmann, N. B., & Koole, S. L. (2007). Dealing with demands: The role of action versus state

orientation. To appear in R. Hoyle (Ed.), The handbook of personality and self-regulation.

Boston, USA: Blackwell Publishing.

Jostmann, N. B., & Koole, S. L. (in press). On the regulation of cognitive control: Action

orientation moderates the impact of situational demands in Stroop interference tasks.

Journal of Experimental Psychology: General.

Jostmann, N. B., Gieselmann, A., & Koole, S. L. (2004). Action orientation moderates the impact

of extensive deliberations on subsequent goal-subgoal conflict resolution in the Tower-of-

Hanoi task. Unpublished dataset, Vrije Universiteit Amsterdam, the Netherlands.

Jostmann, N. B., Koole, S. L., Van der Wulp, N. Y., & Fockenberg, D. A. (2005). Subliminal affect

regulation: The moderating role of action versus state orientation. European Psychologist,


10, 209-217.

Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice-Hall.

Kane, M. J., & Engle, R. W. (2003). Working-memory capacity and the control of attention: The

contributions of goal neglect, response competition, and task set to stroop interference.

Journal of Experimental Psychology: General, 132, 47-70.

Kazén, M., & Kuhl, J. (2005). Intention memory and achievement motivation: Volitional

facilitation and inhibition as a function of affective contents of need-related stimuli. Journal

of Personality and Social Psychology, 89, 426-448.

Kendzierski, D. (1990). Decision making versus decision implementation: An action control

approach to exercise adoption and adherence. Journal of Applied Social Psychology, 20, 27-

45.

Kerns, J. G., Cohen, J. D., MacDonald, A. W., Cho, R. Y., Stenger, V. A., & Carter, C. S. (2004).

Anterior cingulate conflict monitoring and adjustments in control. Science, 303, 1023-1026.

Klein, K., & Boals, A. (2001). The relationship of life event stress and working memory capacity.

Applied Cognitive Psychology, 15, 565-579.

Koole, S. L. (2003). Action versus state orientation among 1,457 Dutch college students.

Unpublished data set. Vrije Universiteit Amsterdam, the Netherlands.

Koole, S. L. (2004). Volitional shielding of the self: Effects of action orientation and external

demand on implicit self-evaluation. Social Cognition, 22, 100-125.

Koole, S. L. (2005). Action versus state orientation, and promotion versus prevention focus.

Unpublished dataset, Vrije Universiteit Amsterdam, the Netherlands.

Koole, S. L., & Jostmann, N. B. (2004). Getting a grip on your feelings: Effects of action

orientation and social demand on intuitive affect regulation. Journal of Personality and

Social Psychology, 87, 974-990.


Koole, S. L., & Kuhl. J. (in press). Dealing with unwanted feelings: The role of affect regulation in

volitional action control. To appear in J. Shah & W. Gardner (Eds.), Handbook of

motivation science. New York: Guilford.

Koole, S. L., Kuhl, J., Jostmann, N. B., & Vohs, K. D. (2005). On the hidden benefits of state

orientation: Can people prosper without efficient affect regulation skills? In: A. Tesser, J.

Woods, & D. A. Stapel (Eds.), Psychological perspectives on the self (pp. 217-244). New

York, NY: Psychology Press.

Koole, S. L., Kuhl. J., Jostmann, N. B., & Finkenauer, C. (2006). Self-regulation in interpersonal

relationships: The case of action versus state orientation. In K. D. Vohs & E. J. Finkel

(Eds.), Self and relationships: Connecting intrapersonal and interpersonal processes (pp.

360-383). New York: Guilford.

Kornblum, S., Hasbroucq, & Osman, A. (1990). Dimensional overlap: Cognitive basis for stimulus-

response compatibility – a model and taxonomy. Psychological Review, 97, 253-270.

Kuhl, J. (1981). Motivational and functional helplessness: The moderating effect of state versus

action orientation. Journal of Personality and Social Psychology, 40, 155-170.

Kuhl, J. (1984). Volitional aspects of achievement motivation and learned helplessness: Toward a

comprehensive theory of action control. In B. A. Maher (Ed.), Progress in experimental

personality research (Vol. 13, pp. 99-171). New York, NY: Academic Press.

Kuhl, J. (1994a). A theory of action and state orientation. In J. Kuhl & J. Beckmann (Eds.), Volition

and personality: Action versus state orientation (pp. 9-46). Göttingen, Germany: Hogrefe &

Huber.

Kuhl, J. (1994b). Action versus state orientation: Psychometric properties of the Action Control

Scale (ACS-90). In J. Kuhl & J. Beckmann (Eds.), Volition and personality: Action versus

state orientation (pp. 47-59). Göttingen, Germany: Hogrefe & Huber.


Kuhl, J. (2000). A functional-design approach to motivation and self-regulation: The dynamics of

personality systems interactions. In M. Boekaerts, P. R. Pintrich, & M. Zeidner, M. (Eds.),

Handbook of self-regulation (pp. 111-169). San Diego, CA: Academic.

Kuhl, J. (2001). Motivation und Persönlichkeit: Interaktionen psychischer Systeme [Motivation and

personality: Personality Systems Interactions]. Göttingen, Germany: Hogrefe.

Kuhl, J., & Beckmann, J. (1994a). Volition and personality: Action versus state orientation.

Göttingen, Germany: Hogrefe & Huber.

Kuhl, J., & Beckmann, J. (1994b). Alienation: Ignoring one’s preferences. In J. Kuhl & J.

Beckmann (Eds.), Volition and personality: Action versus state orientation (pp. 375-389).


Kuhl, J., & Fuhrmann, A. (1998). Decomposing self-regulation and self-control: The volitional

components inventory. In J. Heckhausen & C. S. Dweck (Eds.), Motivation and self-

regulation across the life-span (pp. 15-49). Cambridge: Cambridge University Press.

Kuhl, J., & Goschke, T. (1994a). A theory of action control: Mental subsystems, modes of control,

and volitional conflict-resolution strategies. In J. Kuhl & J. Beckmann (Eds.), Volition and

personality, (pp. 93-124). Göttingen, Germany: Hogrefe & Huber.

Kuhl, J., & Goschke, T. (1994b). State orientation and the activation and retrieval of intentions in

memory. In J. Kuhl & J. Beckmann (Eds.), Volition and personality: Action versus state

orientation (pp. 127-154). Göttingen, Germany: Hogrefe & Huber.

Kuhl, J., & Helle, P. (1986). Motivational and volitional determinants of depression: The

degenerated-intention hypothesis. Journal of Abnormal Behavior, 95, 247-251.

Kuhl, J., & Kazén, M. (1999). Volitional facilitation of difficult intentions: Joint activation of

intention memory and positive affect removes stroop interference. Journal of Experimental



Kuhl, J., & Koole, S. L. (2004). Workings of the will: A functional approach. In: J. Greenberg, S. L.

Koole, & T. Pyszczynski (Eds.), Handbook of experimental existential psychology (pp. 411-

430). New York, NY: Guilford.

Kvavilashvili, L. (1987). Remembering intentions as a distinct form of memory. British Journal of


La Pointe, L. B., & Engle, R. W. (1990). Simple and complex word spans as measures of working

memory capacity. Journal of Experimental Psychology: Learning, Memory, and Cognition,

16, 1118-1133.

Lavie, N., Hirst, A., de Fockert, J. W., & Viding, E. (2004). Load theory of selective attention and

cognitive control. Journal of Experimental Psychology: General, 133, 339-354.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping. New York, NY: Springer.

Lindgren, A. (1985). Ronia, the robber’s daughter. New York, NY: Penguin.

Logan, G. D., & Zbrodoff, N. J. (1979). When it helps to be misled: Facilitative effects of

increasing the frequency of conflicting stimuli in a Stroop-like task. Memory and Cognition,

7, 166-174.

Long, D. L., & Prat, C. S. (2002). Working memory and Stroop interference: An individual

difference investigation. Memory and Cognition, 30, 294-301.

Lorist, M. M., Boksem, M. A. S., & Ridderinkhof, K. R. (2005). Impaired cognitive control and

reduced cingulate activity during mental fatigue. Cognitive Brain Research, 24, 199-205.

Lorist, M. M., Klein, M., Nieuwenhuis, S., De Jong, R., Mulder, G., & Meijman, T. F. (2000).

Mental fatigue and task control: Planning and preparation. Psychophysiology, 37, 614-625.

Lundqvist, D., Esteves, F., & Öhman, A. (1999). The face of wrath: Critical features for conveying

facial threat. Cognition and Emotion, 13, 691-711.

Luria, A. R. (1966). Higher cortical functions in man. London, UK: Tavistock.


Luria, A. R., & Tsvetkova, L. D. (1964). The programming of constructive ability in local brain

injuries. Neuropsychologia, 2, 95-108.

MacDonald, A. W., Cohen, J. D., Stenger, V. A., & Carter, C. S. (2000). Dissociating the role of the

dorsolateral prefrontal and anterior cingulate cortex in cognitive control. Science, 288, 1835-

1838.

MacLeod, C. M. (1991). Half a century of research on the Stroop effect: An integrative review.

Psychological Bulletin, 109, 163-203.

Markus, H., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and

motivation. Psychological Review, 98, 224-253.

Marsh, R. L., Hicks, J. L., & Bink, M. L. (1998). Activation of completed, uncompleted, and

partially completed intentions. Journal of Experimental Psychology: Learning, Memory,

and Cognition, 24, 350-361.

Martijn, C., Tenbült, P., Merckelbach, H., Dreezens, E., & de Vries, N. (2002). Getting a grip on

ourselves: Challenging expectancies about loss of energy after self-control. Social

Cognition, 20, 441-460.

Martin, M., Kliegel, M., & McDaniel, M. A. (2003). The involvement of executive functions in

prospective memory performance of adults. International Journal of Psychology, 38, 195-

206.

Mayr, U., Awh, E., & Laurey, P. (2003). Conflict adaptation and effects in the absence of executive

control. Nature Neuroscience, 6, 450-452.

McClelland, D. C., Koestner, R., & Weinberger, J. (1989). How do self-attributed and implicit

motives differ? Psychological Review, 96, 690-702.

Meichenbaum, D. H., & Deffenbacher, J. L. (1988). Stress inoculation training. The Counseling



Miller, G. A., Galanter, E., & Pribram, K. (1960). Plans and the structure of behavior. New York,

NY: Holt, Rinehart, & Winston.

Mischel, W., Cantor, N., & Feldman, S. (1996). Principles of self-regulation: The nature of

willpower and self-control. In E. T. Higgins, & A. W. Kruglanski (Eds.), Social psychology:

Handbook of basic principles (pp. 329-360). New York, NY: Guilford.

Miyake, A., & Shah P. (1999). Models of working memory: Mechanisms of active maintenance and

executive control. New York, NY: Cambridge University Press.

Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., & Howerter, A. (2000). The unity and

diversity of executive functions and their contributions to complex “frontal lobe” tasks: A

latent variable analysis. Cognitive Psychology, 41, 49-100.

Monsell, S. (1996). Control of mental processes. In: V. Bruce (Ed.). Unsolved mysteries of the

mind: Tutorial essays in cognition. (pp. 93-148). Oxford, UK: Erlbaum.

Moors, A., & De Houwer, J. (2006). Automaticity: A theoretical and conceptual analysis.

Psychological Bulletin, 132, 297-326.

Moskowitz, G. B., Li, P., & Kirk, E. R. (2004). The implicit volition model: On the preconscious

regulation of temporarily adopted goals. In M. Zanna (Ed.), Advances in experimental social

psychology (Volume 34, pp. 317-414). San Diego, CA: Academic Press.

Muraven, M., & Baumeister, R. F. (2000). Self-regulation and depletion of limited resources: Does

self-control resemble a muscle? Psychological Bulletin, 126, 247-259.

Muraven, M., Baumeister, R. F., & Tice, D. M. (1999). Longitudinal improvement of self-

regulation through practice: Building self-control strength through repeated exercise.

Journal of Social Psychology, 139, 446-457.

Muraven, M., Tice, D. M., & Baumeister, R. F. (1998). Self-control as limited resource: Regulatory

depletion patterns. Journal of Personality and Social Psychology, 74, 774-789.


Murphy, S. T., & Zajonc R. (1993). Affect, cognition, and awareness: Affective priming with

optimal and suboptimal stimulus exposures. Journal of Personality and Social Psychology,

64, 723-739.

Neumann, O. (1987). Beyond capacity: A functional view of attention. In H. Heuer & A. F. Sanders

(Eds.), Perspectives on perception and action (pp.361-394). Hillsdale, NJ: Erlbaum.

Neumann, R., Förster, J., & Strack, F. (2003). Motor-compatibility: The bidirectional link between

behavior and evaluation. In J. Musch & K. C. Klauer (Eds.), The psychology of evaluation:

Affective processes in cognition and emotion (pp. 371-391). London: Erlbaum.

Nigg, J. T. (2001). Is ADHD a disinhibitory disorder? Psychological Bulletin, 127, 571-598.

Norman, D. A., & Shallice, T. (1986). Attention to action: Willed and automatic control of

behavior. In R. J. Davidson, G. E. Schwartz, & D. Shapiro (Eds.), Consciousness and self-

regulation: Advances in research and theory (Vol. 4, pp. 1-18). New York: Plenum.

Öhman, A., Lundqvist, D., Esteves, F. (2001). The face in the crowd revisited: A threat advantage

with schematic stimuli. Journal of Personality and Social Psychology, 80, 381-396.

Olivers, C. N. L., & Nieuwenhuis, S. (2006). The beneficial effects of additional task load, positive

affect, and instruction on the attentional blink. Journal of Experimental Psychology: Human

Perception and Performance, 32, 364-379.

Olvermann, R., Metz-Göckel, H., Hannover, B., & Pöhlmann, C. (2004). Motivinhalte und

Handlungs-versus Lageorientierung bei independenten oder interdependenten Personen

[Contents of Motives and Action Versus State Orientation in Independent or Interdependent

Persons]. Zeitschrift für Differentielle und Diagnostische Psychologie, 25, 87-103.

Palfai, T. P. (2002). Action-state orientation and the self-regulation of eating behavior. Eating

Behaviors, 3, 249-259.


Palfai, T. P., McNally, A. M., & Roy, M. (2002). Volition and alcohol-risk reduction: The role of

action orientation in the reduction of alcohol-related harm among college student drinkers.

Addictive Behaviors, 27, 309-317.

Pashler, H. (1994). Dual-task interference in simple tasks: Data and theory. Psychological Bulletin,

116, 220-244.

Pollack, I. (1970). A nonparametric procedure for evaluation of true and false positives. Behavior

Research Methods and Instrumentation, 2, 155-156.

Posner, M. I., & Snyder, C. R. R. (1975). Attention and cognitive control. In R. L. Solso (Ed.),

Information processing and cognition (pp. 55-85). Hillsdale, NJ: Erlbaum.

Ratcliff, R. (1993). Methods for dealing with reaction time outliers. Psychological Bulletin, 114,

510-532.

Reason, J., & Mycielska, K. (1982). Absent-minded? The psychology of mental lapses and everyday

errors. Englewood Cliffs, NJ: Prentice Hall.

Rholes, W. S., Michas, L., & Shroff, J. (1989). Action control as a vulnerability factor in dysphoria.

Cognitive Therapy and Research, 13, 263-274.

Ridderinkhof, K. R., Ullsperger, M., Crone, E. A., & Nieuwenhuis, S. (2004). The role of the

medial frontal cortex in cognitive control. Science, 306, 443-447.

Roe, W. T., Wilsoncroft, W. E., & Griffiths, R. S. (1980). Effects of motor and verbal practice on

the Stroop task. Perceptual and Motor Skills, 50, 647-650.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University

Press.

Ryan, R. M., & Deci, E. L. (2004). Autonomy is no illusion: Self-determination theory and the

empirical study of authenticity, awareness, and will. In: J. Greenberg, S. L. Koole, & T.

Pyszczynski (Eds.), Handbook of experimental existential psychology (pp. 449-479). New

York, NY: Guilford.


Scheier, M. F., & Carver, C. S. (1985). Optimism, coping, and health: Assessment and implications

of generalized outcome expectancies. Health Psychology, 4, 219-247.

Schmader, T., & Johns, M. (2003). Converging evidence that stereotype threat reduces working

memory capacity. Journal of Personality and Social Psychology, 85, 440-452.

Schmeichel, B. J., Vohs, K. D., & Baumeister, R. F. (2003). Intellectual performance and ego

depletion: Role of the self in logical reasoning and other information processing. Journal of


Schneider, W., R Shiffrin, R. M. (1977). Controlled and automatic human information processing:

I. Detection, search, and attention. Psychological Review, 84, 1-66.

Schönpflug, W. (1983). Coping efficiency and situational demands. In G. R. J. Hockey (Ed.), Stress

and Fatigue in Human performance, (pp. 299-330). Norwich, UK: Wiley.

Schulte, D., Hartung, J., & Wilke, F. (1997). Handlungskontrolle der Angstbewältigung: Was

macht Reizkonfrontationsverfahren so effektiv? [Action control in treatment of anxiety

disorders: What makes exposure so effective?]. Zeitschrift für Klinische Psychologie, 26,

118-128.

Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being:

Informative and directive functions of affective states. Journal of Personality and Social


Segerstrom, S. C., & Miller, G. E. (2004). Psychological stress and the human immune system: A

meta-analytic study of 30 years of inquiry. Psychological Bulletin, 130, 601-630.

Seligman, M. E. P. (1975). Helplessness. On depression, development and death. San Francisco,

CA: Freeman.

Shah, J. Y. (2003). The motivational looking glass: How significant others implicitly affect goal

appraisals. Journal of Personality and Social Psychology, 85, 424-439.

Shah, J. Y., & Higgins, E. T. (1997). Expectancy X value effects: Regulatory focus as determinant


of magnitude and direction. Journal of Personality and Social Psychology, 73, 447-458.

Shah, J. Y., Kruglanski, A. W., & Friedman, R. (2003). Goal systems theory: Integrating the

cognitive and motivational aspects of self-regulation. In S. J. Spencer, & S. Fein (Eds.)

Motivated social perception: The Ontario symposium on personality and social psychology

(pp. 247-275). Mahwah, NJ: Erlbaum.

Simon, H. (1967). Motivational and emotional controls of cognition. Psychological Review, 74, 29-

39.

Stapel, D. A. (2002). Affective reactions to subliminally presented schematic drawings of happy

and sad faces. Unpublished dataset. University of Groningen, the Netherlands.

Stapel, D. A., Koomen, W., Ruys, K. I. (2002). The effects of diffuse and distinct affect. Journal of


Stepper, S., & Strack, F. (1993). Propioceptive determinants of emotional and non-emotional

feelings. Journal of Personality and Social Psychology, 64, 211-220.

Stiensmeier-Pelster, J. (1994). Choice of decision-making strategies and action versus state

orientation. In J. Kuhl & J. Beckmann (Eds.), Volition and personality, (pp. 167-176).


Stiensmeier-Pelster, J., John, M., Stulik, A., & Schürmann, M. (1989). Die Wahl von

Entscheidungsstrategien: Der Einfluß von Handlungs- und Lageorientierung und die

Bedeutung psychologischer Kosten. [The choice of decision-making strategies: The effects

of action vs. state orientation and the role of psychological costs.] Zeitschrift für

Experimentelle und Angewandte Psychologie, 36, 292-310.

Stroop, J. R. (1935). Studies of interference in serial verbal reactions. Journal of Experimental



Tesser, A., Millar, M., & Moore, J. (1988). Some affective consequences of social comparison and

reflection processes: The pain and pleasure of being close. Journal of Personality and Social


Tiffany, S. T. (1990). A cognitive model of drug urges and drug-use behavior: Role of automatic

and nonautomatic processes. Psychological Review, 97, 147-168.

Turner, M. L., & Engle, R. W. (1989). Is working memory capacity task dependent? Journal of

Memory and Language, 28, 127-154.

Van der Linden, D., Frese, M., & Meijman, T. F. (2003). Mental fatigue and the control of

cognitive processes: Effects on perseveration and planning. Acta Psychologica, 113, 454-65.

Van Stekelenburg, J. (2006). Promoting or preventing social change. Instrumentality, identity,

ideology and group-based anger as motives of protest participation. Doctoral dissertation,

Vrije Universiteit Amsterdam, the Netherlands.

Vohs, K. D., & Finkel, E. J. (2006). Self and relationships: Connecting intrapersonal and

interpersonal processes. New York, NY: Guilford.

Wegner, D. M. (2002). The illusion of conscious will. Cambridge, MA: MIT.

Wegner, D. M., & Bargh, J. A. (1998). Control and automaticity in social life. In

D. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (pp. 446-

496). New York, NY: McGraw-Hill.

Wilson, T. D., & Dunn, E. (2004). Self-knowledge: Its limits, value, and potential for improvement.

Annual Review of Psychology, 55, 493-518.

Winkielman, P., Berridge, K .C., & Wilbarger, J. L. (2005). Unconscious affective reactions to

masked happy versus angry faces influence consumption behavior and judgments of value.

Personality and Social Psychology Bulletin, 31, 121-135.

Winkielman, P., Zajonc, R. B., & Schwarz, N. (1997). Subliminal affective priming resists

attributional interventions. Cognition and Emotion, 11, 433-465.


Zajonc, R. B. (2000). Feeling and thinking: Closing the debate over the independence of affect. In J.

P. Forgas (Ed.), Feeling and thinking: The role of affect in social cognition (pp. 31-58).

New York: Cambridge University Press.

Zeidner, M., & Endler, N. S. (1996). Handbook of coping. New York, NY: Wiley.

Zeigarnik, B. (1927). Über das Behalten von erledigten und unerledigten Handlungen [On

remembering finished and unfinished activities]. Psychologische Forschung, 9, 1-85.

Zwaan, R. A. & Truitt, T. P. (1998). Smoking urges affect language processing. Experimental and

Clinical Psychopharmacology, 6, 325-330.

Documents

When The Going Gets Tough… How Action Versus State ... dissertation.pdfWhen The Going Gets Tough… How Action Versus State Orientation Moderates The Impact Of Situational Demands