26
This manuscript was published as: Ruch, W., Harzer, C., Proyer, R. T., Park, N., & Peterson, C. (2010). Ways to happiness in German-speaking countries: The adaptation of the German version of the Orientations to Happiness Questionnaire in paper-pencil and internet samples. European Journal of Psychological Assessment, 26, 227- 234.

Ways to Happiness in German-Speaking Countries: The Adaptation of the German Version of the Orientations to Happiness Questionnaire in Paper-Pencil and Internet Samples

Embed Size (px)

Citation preview

This manuscript was published as: Ruch, W., Harzer, C., Proyer, R. T., Park, N., & Peterson, C. (2010). Ways to happiness in German-speaking countries: The adaptation of the German version of the Orientations to Happiness Questionnaire in paper-pencil and internet samples. European Journal of Psychological Assessment, 26, 227-234.

1

Running head: ORIENTATIONS TO HAPPINESS

Ways to Happiness in German-Speaking Countries: The Adaptation of the German Version of

the Orientations to Happiness Questionnaire in Paper-Pencil and Internet Samples

Willibald Ruch

Claudia Harzer

René T. Proyer

University of Zurich, Zurich, Switzerland

Nansook Park

Christopher Peterson

University of Michigan, Ann Arbor, MI, USA

Send proofs and reprints to:

Willibald Ruch, Section on Personality and Assessment, Department of Psychology,

University of Zurich, Binzmuehlestrasse 14/7, 8050 Zurich, Switzerland.

Phone: office: +41 44 635 75 20; Fax +41 44 635 75 29; e-mail: [email protected]

Orientations to happiness

2

Summary

Peterson, Park, and Seligman (2005) developed the Orientations to Happiness (OTH)

questionnaire to measure three routes to happiness: life of pleasure (hedonism), life of

engagement (flow) and life of meaning (eudemonia). The questionnaire was translated into

German in several steps (independent translations by five experts; creating the initial version

by committee approach; retranslation; modification; and final version). Data were collected in

paper-pencil (N = 1,152) and Internet samples (N = 4,174). The OTH scales showed

satisfactory internal consistencies (.63 ≤ α ≤ .76) and stability across six months (all ≥ .63).

The factorial structure of the German OTH was analyzed (RMSEA ≤ .074 and SRMR ≤ .043

for the three-factor solution) in both samples. Tucker's Phi coefficients for factorial

congruence between the two samples were .99. The three scales were positively

intercorrelated. The endorsement of the life of pleasure was higher in younger, unmarried, and

not religious participants. The life of meaning was more pronounced among the religious

participants. There was a good convergence (all > .49) between homologous scale of the self-

and peer-form, and the OTH predicted behavior in prototypical pleasure, engagement and

meaning situations. Most importantly, high scores in each of the orientations to happiness

reported higher degrees of overall satisfaction with life.

Keywords: Orientations to Happiness, positive psychology, pleasure, engagement, meaning

Orientations to happiness

3

Ways to Happiness in German-Speaking Countries: The Adaptation of the German Version of

the Orientations to Happiness Questionnaire in Paper-Pencil and Internet Samples

Positive psychology studies positive subjective experiences, positive individual traits, and

positive institutions (Seligman & Csikszentmihalyi, 2000). Among its prime topics is the

scientific study of subjective well-being and happiness. Seligman (2002) described three

different ways of obtaining happiness: the life of pleasure; the life of engagement; and the life

of meaning. The pleasurable life reflects the route to happiness through hedonism by

“maximizing pleasure and minimizing pain” (Peterson, Park, & Seligman, 2005, p.25). The

second orientation to happiness is the engaged life, which is based on Csikszentmihalyi’s

(1990) concept of flow. Csikszentmihalyi found that people who are absorbed by an engaging

activity (i.e., experienced flow) are more satisfied, and consequently happier. The life of

meaning reflects the concept of eudaimonia (“being true to one’s inner self/demon”; Peterson

et al., 2005, p.26). It reflects the route of achieving happiness by identifying one’s virtues,

cultivating them, and living in accordance to them to achieve a higher purpose. According to

Seligman (2002), these orientations are not mutually exclusive but can be pursued

simultaneously. He also suggested that they are suitable for enhancing happiness.

Measuring the three orientations to happiness. Peterson et al. (2005) developed a self-

report questionnaire (Orientations to Happiness; OTH) that measures endorsement of the life

of pleasure (P), engagement (E), and meaning (M). The scale yielded high internal

consistencies and in factor analyses (Varimax rotation), Peterson et al. (2005) found support

for the expected three-factor solution. Each item loaded substantially on one factor but several

items had a second loading on other factors as well. In two studies, the correlations between P

and M were lowest (r = .17 and r = .18), whereas the one between E and M was the highest (r

Orientations to happiness

4

= .42 and r = .46). The correlations between P and E were r = .31 and r = .32. The correlations

showed that the three orientations are related but not interchangeable (Peterson et al., 2005).

The life of pleasure was slightly associated with lower age, lower education, and not

being married. E and M were independent of socio-demographic variables. Gender was not

related to any of the three scales. Most importantly, each of the three orientations correlated

positively with life satisfaction (r = .17 to r = .30; Peterson et al., 2005). E and M were more

robustly related to life satisfaction than P.

Aims of the present study. The present paper describes the adaptation of the German

version of the OTH aiming at obtaining a highly similar version compared to the original

English questionnaire. Translation-retranslation procedure, descriptive statistics, factorial

structure, reliability (i.e., internal consistency and retest reliability), and validity (i.e.,

convergent and discriminant, and criterion validity) will be considered. Furthermore, the

pattern of intercorrelations and correlation with satisfaction with life (i.e., criterion validity)

should be replicated. Finally, the relations to demographic variables should be comparable to

the results found for the US version.

The rationale for examining the criterion validity of the OTH was based on two

different approaches. Firstly and most importantly, it was argued that a measure that describes

orientations to happiness should be related to subjective measures of overall well-being.

These relations were described in studies using the American OTH (Peterson et al., 2005;

Peterson, Ruch, Beermann, Park, & Seligman, 2007) and they should be replicable with the

present adaptation if the scales are equivalent. It is expected that all three scales of the OTH

will be modestly positively related to measures of overall satisfaction with life, with E

yielding the highest and P the lowest coefficient. The second approach was aimed at testing

convergent and discriminant validity when predicting (self-reports of) behaviors relating to

these three orientations. It was expected, for example, that someone pursuing a life of

Orientations to happiness

5

engagement will spend more time planning a prototypical “engagement” activity, spend more

time indulging in it, and that one ideally would want to spend most of ones time with these

activities. Therefore, we created three different scenarios (i.e., prototypical for each of the

three orientations). The first scenario dealt with a pleasurable activity at leisure time, the

second one with an engaging activity at work, and the third one with a meaningful activity in

a family-context. Additionally and for the first time, a peer-rating form of the OTH was used

to test its convergence with the self-ratings. A high convergent validity supports the construct

validity of the German version.

In the present study, religiousness was taken into account as an additional demographic

variable. It was expected that participants who actively pursue religion would score higher on

M and lower on P than non-religious individuals. The item contents in the life of meaning

scale reflect a responsibility for the world or a sense of a life that has a higher purpose. Thus,

these contents are similar to the emphases of many religions. The pleasurable life, on the

other hand, might be interpreted as a weakness from a religious point of view, and therefore,

negative relations were expected.

The psychometric properties, intercorrelations, and factorial structure were examined

for data collected in two different ways: in a traditional paper-pencil form and in an Internet

survey. If the results converge well, this is evidence for the usefulness of the measure as well.

Materials and Methods

Research participants

Sample 1 (Paper-pencil). Sample 1 consisted of 1,212 German-speaking adult volunteers

from Austria, Germany, and Switzerland (848 women, 364 men). Their age ranged from 18 to

88 years, with a mean of 36.23 years (SD = 13.93). N = 425 were students, and N = 788 was a

mixed sample of adults. About 88% of the respondents provided information about their

Orientations to happiness

6

education: 26% of them had an apprenticeship, 23% baccalaureate, and 50% a University

degree. With respect to marital status, 87% of the respondents provided information; 43% of

them were married or lived as married, and 57% were single. 54 % provided information

about their religion; 70% were Christians, 21% indicated having no religion, and the others

were followers of Free Church, Islam, and Buddhism.

Sample 2 (Internet). Sample 2 included N = 4,174 German-speaking adults from

Austria, Germany, and Switzerland (2,662 women and 1,512 men) from 18 to 86 years of age

(M = 41.55, SD = 12.98). The educational level was compulsory education for 9% of the

participants, 35% had an apprenticeship, 13% baccalaureate, and 44% a University degree.

About 52% were married or lived as married, and 48% of respondents were single. Being

asked for their religion, 67% specified to be Christians, 24% indicated no religion, and the

others were followers of Free Churches, Judaism, Hinduism, and Buddhism.

Validation sample. The sample consisted of N = 201 participants. Their age ranged

from 18 to 77 years (M = 44.76, SD = 13.54). 70 were males and 131 were females. About

one third (32.84%) were married and about one fourth (23.38%) were not in a relationship.

Instruments

The Orientation to Happiness measure (OTH; Peterson et al., 2005) is an 18-item self-

report questionnaire for the subjective assessment of life of pleasure, life of engagement, and

life of meaning (six items each). It utilizes a 5-point Likert-scale (“1 = very much unlike me”

through “5 = very much like me”). A sample item for life of engagement is “I am always very

absorbed in what I do”, for life of pleasure “Life is too short to postpone the pleasures it can

provide”, and for life of meaning a sample item is “I have a responsibility to make the world a

better place”. A peer-report form of the OTH was created which is identical to the self-report

form except the items are formulated in third person singular.

Orientations to happiness

7

The Satisfaction with Life Scale (SWLS; Diener, Emmons, Larsen, & Griffin, 1985) is a

five-item measure for assessing the satisfaction with life using a 7-point answer format (from

1 = “strongly disagree” through 7 = “strongly agree”). A sample item is “The conditions of

my life are excellent”. The SWLS yielded a high internal consistency in the present sample (α

= .88). The SWLS is widely used and showed good psychometric properties in various studies

(e.g., Diener 1994).

The Orientations to Happiness-Situations Rating Form consisted of short descriptions of

three situations that were set together to subjectively assess behavior in leisure time

(pleasure), at work (engagement), and in family-related situations (meaning). For example, in

the leisure condition the task of the subjects was to imagine activities that they enjoy in their

leisure time such as hobbies, or activities that they pursue to grant themselves. Thus, they

should think of situations in which they primarily have fun. The participants then answered

three questions related to these situations; i.e., (a) how much time they spend for planning

such activities; (b) how much time they spend pursuing such activities; and (c) whether they

would primarily pursue such activities under ideal conditions. Answers were given on a 5-

point rating form from 1 = “very much like me” to 5 = “very much unlike me”.

Procedure

Five psychologists translated the OTH, and the initial version of the German OTH was

created by committee approach (Butcher & Pancheri, 1976). A bilingual psychologist

retranslated the first version of the German OTH. A few modifications were made to that

initial version, and German items were checked for understandability. The items of the final

German version of the OTH are shown in the Appendix A.

Sample 1 (Paper-pencil) was recruited in several different ways to obtain a heterogeneous

sample. Students were tested at the university and the other participants (public advertised

participants, volunteers from residential homes) completed the questionnaire at home and sent

Orientations to happiness

8

it back by mail. Respondents were not paid for participating, but received standardized

feedback about their individual scores relative to other respondents upon request.

Sample 2 (Internet). Participants completed the OTH among other instruments on a well

established website (www.charakterstaerken.org; hosted by the Section on Personality and

Assessment of the University of Zurich) for research purposes between January 2006 and

June 2007. The website was promoted by different means, such as press coverage (e.g.,

newspaper and several magazines) and contacting special groups (e.g., artists, police,

theologians) in order to ascertain heterogeneity of the sample. The volunteers registered on

the website from their personal computers and completed the questionnaire there.

Respondents were not paid for participating, but were provided an automatically generated

feedback of individual results. Testing via the Internet has been criticized in different

occasions (e.g., for sample biases). However, there are standards for the implementation of

Internet-delivered-testing (Coyne & Bartram, 2006) that facilitate this way of data collection.

Furthermore, there is empirical evidence that data collected via the Internet leads to similar

findings as more traditional paper-pencil methods (Gosling, Vazire, Srivastava, & John,

2004). A sub sample additionally completed the SWLS (N = 303, 18 to 81 years; M = 39.61,

SD = 13.02, 132 males and 171 females).

The participants of the validation sample completed the OTH and the Orientations to

Happiness-Situations Rating Form along with other questionnaires as part of a study on

positive psychology. The questionnaires were mailed to them and they completed all

questionnaires in paper-pencil form. They were not paid for their services and received a

standardized feedback on their results after the completion of the study.

Orientations to happiness

9

Analyses of data

Analyzing factorial structure, we tested the model fit for a one- to four-factor solution

of the data by using Mplus (Version 5; Muthén & Muthen, 2007). The WLSMV estimator

was chosen because of its applicability for categorical endogenous data at the item level (e.g.,

Hancock & Mueller, 2006). Therefore, analyses concerning the factorial structure were based

on polychoric correlations. Criteria for the evaluation of the models were the p-value of chi-

square (χ2; Hair, Anderson, Tatham, & Black, 2006), the Standard Root Mean Square

Residual and Root Mean Square Error of Approximation (SRMR and RMSEA; Hu & Bentler,

1998). A non-significant p-value of chi-square (χ2) indicates a good fit. But the χ2-value is not

independent of sample size (Hair et al., 2006). Therefore, due to the large sample sizes in both

samples (Paper-pencil and Internet) a significant value was to be expected. For the SRMR and

the RMSEA values equal or lower .08 can be interpreted as a good fit.

A one-dimensional solution would indicate that there is a general orientation to

happiness factor that comprises pleasurable, engaged, and meaningful aspects. A two-factor

solution might provide support for the philosophical approach. In this case, it might be that

hedonism and eudemonia fully explain the engaging life. A three-factorial solution would

replicate prior empirical findings with the American OTH and the theoretical assumptions.

The American authors of the OTH based their rationale for a three-factor solution on the

notion that both psychological but also other resources (e.g., philosophical) suggested that

there are three routes to a good life. While two of them were already well described (i.e.,

hedonism and eudemonia) the third one was added based on psychological premises (i.e., that

flow-experiences contribute to the good life). In this case the German OTH would cover all

three orientations as well. It was expected that this solution would yield the best fit-indices.

However, it seemed possible that a four-factor solution also might apply. For example, four

Orientations to happiness

10

factors might appear if the life of meaning splits up into two factors describing meaning

endorsed by a person individually and meaning in relation to the whole society.

The similarity of the factors in the paper-pencil and in the Internet sample was

examined by means of the Tucker’s Phi coefficient (Tucker, 1951). Coefficients equal to or

higher .90 are considered as adequate while coefficients ≥ .95 are considered to be excellent.

Results

Factor structure1. For the examination of the factorial structure of the OTH, we tested

the fit of one to four factor models to the data. For replicating the findings with the American

OTH we used the same specifications for the analyses as in the paper describing the

construction of the scale. However, we used Mplus (Version 5) for the exploratory factor

analyses as it provides estimators for categorical data (WLSMV). The first six Eigenvalues

were 4.27, 2.64, 1.57, 1.02, 0.94, and .90 for the Online and 4.35, 2.44, 1.55, 1.08, 1.01, and

.92 for the Paper-Pencil data. Table 1 shows the goodness of fit indicators of the four different

factor models in the two samples.

Insert Table 1 about here

1 The factorial structure of the German OTH might also be studied by means of Confirmatory Factor Analysis (CFA). This was initially pursued but later this strategy was abandoned. More precisely, different models were tested. The one with the most restrictive assumptions assumed uncorrelated factors and six items measuring one factor solely without secondary loadings. This analysis yielded unsatisfying fits in both, the paper-pencil and the Internet sample (χ2[47] = 1,652.08, p < .001 and χ2[52] = 5,316.52, p < .001; CFI = .67 and CFI = .69 for the paper-pencil and the Internet sample, respectively). Less restrictive models that fit the theoretical assumptions and this area of research better yielded slightly better but still not satisfying results. For example, a factorial model assuming intercorrelated factors and six items measuring one factor solely without secondary loadings was tested (χ2[77] = 837.92, p < .001 and χ2[82] = 3,672.66, p < .001; CFI = .82 and CFI = .79 for the paper-pencil and the Internet sample, respectively). Finally, it was decided to keep the analyses strictly parallel to the ones used for the original scale.

Orientations to happiness

11

Table 1 shows that the models assuming three and four factors fitted well (all indices

indicated good fit) in both samples. Due to the high number of participants, the χ2-value

(CMIN) demonstrated statistical significance. The increase of goodness of fit was higher from

the one- to the two- (e.g., ΔRMSEA = .062, ΔSRMR = .055) and from the two- to the three-

factorial solutions (e.g., ΔRMSEA = .032, ΔSRMR = .024) than from the three- to the four-

factorial (e.g., ΔRMSEA = .007, ΔSRMR = .006) solutions. This pattern was found in both

samples. The more parsimonious three-factor model yielded highly similar fit indices

compared to the more sophisticated four-factor model. Therefore, for practical purposes, the

three-factor solution was favored (cf. Bollen, 1989). Additionally, across both samples the

scree test suggested the retention of three factors rather than four. Table 2 shows the factor

loadings for the three-factor models rotated to the VARIMAX-criterion2 in the paper-pencil

and the Internet samples.

Insert Table 2 about here

Table 2 shows that the main loadings of the paper-pencil sample ranged from .40 to .71

(median = .59), from .33 to .76 (median = .38), and from .36 to .72 (median = .64) for P, E

and M, respectively. As in the American OTH, the items 4,5, 8, and 9 had secondary loadings.

2 In the analyses conducted so far the factors were always rotated according to the VARIMAX-criterion. We did stick to this decision too, to create optimal conditions for a successful replication of the findings by Peterson et al. (2005). However, as the scales are intercorrelated an oblique rotation might also be indicated. Therefore, we repeated the analyses (keeping all parameters equal) but rotated the data according to the OBLIMIN-criterion. Overall, the pattern of loadings for both samples for a three-factor solution was highly similar to the ones of the VARIMAX-rotation. The medians of the loadings in the paper pencil sample were .58 (range .39 to 74), .34 (.29 to .77), and .66 (.38 to .74) for pleasure, engagement, and meaning, respectively; and .61 (range .35 to 73), .36 (.32 to .76), and .58 (.34 to .81) in the Internet sample. As in the VARIMAX-solution, the items 4, 5, 8, and 9 yielded loadings on secondary factors. Again, item 9 demonstrated higher loadings on the non-target factor. Factor intercorrelations were .38/.35 between engagement and meaning, .24/.32 between engagement and pleasure, and .08/.13 between meaning and pleasure for the Internet/paper pencil data.

Orientations to happiness

12

However, in the German OTH item 9 yielded a higher loading on pleasure than on

engagement.

Similar loading patterns were found in the Internet sample with .61, .39, and .59 as

medians of the main loadings for P, E, and M, respectively. Tucker's Phi coefficient was

computed as a measure of the congruence between the paper-pencil and Internet samples. The

coefficients for the three factors were excellent and indicated that the German OTH measured

equivalent dimensions in the two samples. Tucker's Phi indicated identical (.99) patterns of

factor loadings for the three orientations to happiness.

Descriptive statistics for the German version of the OTH. Mean scores and standard

deviations were computed for each of the three scales. Skewness and kurtosis indicated that

the scales were normally distributed. Furthermore, the internal consistencies, the corrected-

item total correlations, and data on the stability of the OTH were analyzed. Finally, the

intercorrelations and correlations with, age, gender, educational level, and practicing one’s

religion (yes, no, no answer) were computed (see Table 3).

Insert Table 3 about here

Table 3 shows that the corrected item-total correlations were in an acceptable range.

The internal consistencies exceeded the scale intercorrelations by far. The alpha coefficients

for P and M were ≥ .73 in both versions. E had the lowest reliability in both the Paper-pencil

(α = .64) and the Internet sample (α = .63). In several studies, the OTH was administered

more than once allowing for an estimation of the retest reliability. The test-retest intervals

were 3 month (N = 113 from 19 to 77 years, M = 45.24, SD = 14.19; 39 males, 74 females)

and six month (N = 96 from 22 to 77 years; M = 44.46, SD = 12.71; 34 males, 62 females).

The test-retest correlations shown in Table 2 were about the size of the internal consistencies

suggesting a high stability of the German OTH-scales for short time intervals.

Orientations to happiness

13

The scale means were significantly higher in the Internet sample than in the paper-

pencil sample. The effect-sizes of these differences were low and ranged between η2 = .002

for P and η2 = .024 for E. The rank order of the scale means was highly similar in both

samples (highest in P and lowest in M). The standard deviations of the scales in the paper-

pencil and the Internet versions were comparable; they were smallest for E in both samples.

The intercorrelations among the OTH scales (controlled for age, gender, and education)

were moderately positive in the two samples, and they yielded a highly similar pattern in both

samples. However, the intercorrelations were higher in the paper-pencil sample than in the

Internet sample. In both samples, the highest correlation was between P and E, and the lowest

between P and M. Furthermore, the correlations with demographic variables were generally

low, yet statistically significant in a few cases due to the high number of participants.

However, a few meaningful relations were revealed. For example, P was negatively related to

age indicating that younger participants described higher levels of hedonism. Similarly,

unmarried and non-religious participants also had higher endorsements to P. Religious people

described themselves with higher expressions of M. Men and women were similar in their

expressions in all of the three orientations3.

Convergence between self-report and peer-report forms of the OTH. In a study

involving 127 psychology students (M = 25.45, SD = 6.46, 19 to 54 years, 24 males and 103

females) self- and peer report data (two peers) on the OTH were collected (using a paper-

pencil administration). The correlations between the two data sources are shown in Table 4.

Insert Table 4 about here

3 The correlation coefficients were highly similar when computing partial correlations controlling for the other demographic variables. The only exception was that the marriage-pleasure relationship faded after controlling for age, sex, educational level, and religiousness.

Orientations to happiness

14

Table 4 shows that self- and peer reports converged well with the highest correlations in

the diagonal; the mean of the convergent validity coefficients was r = .50. Like in the self-

reports the correlations between P and E, and between E and M were higher than the

correlations between P and M in the heteromethod blocks. Interestingly, the heteromethod

correlations between P and M were slightly negative (albeit non-significant). However, in the

mono-method correlations (not shown here) they were virtually zero and so one cannot say

that this is really reflected method variance.

Prediction of satisfaction with life. All three orientations to happiness were positively

correlated with satisfaction with life (M = 4.77, SD = 1.26). The correlation coefficients were

r = .26, r = .29, and r = .24 (all p < .01; N = 303) for pleasure, engagement, and meaning,

respectively. Stepwise multiple regressions with the OTH-scales as predictors and the total

score of the SWLS were performed. All three scales entered the equation and yielded a

multiple correlation coefficient of .36 (F[1,302] = 15.18; p < .01).

Predicting prototypical activities. The OTH had similar psychometric properties in the

validation sample as in the two samples used for determining the factorial structure. The

Orientations to Happiness-Situations Rating Form provided information on the amount of

time participants spent on planning and pursuing pleasurable activities in their leisure time,

engaged activities at work, and meaningful activities in family related situations.

Additionally, scores for their wish to ideally spend time with those activities were collected.

Table 5 contains the zero-order correlations with the OTH, as well as multiple correlations

with these three items as predictors and the OTH scales as a criterion.

Insert Table 5 about here

Table 5 shows that the highest correlations were found for the main axis of the matrix.

This indicated that higher scores in pleasure were related to more planning and pursuing of

Orientations to happiness

15

pleasure activities and to the wish to do so primarily often under ideal circumstances. The

same patterns applied to the work-engagement and family-meaning situations. However, the

engaged life was also related to planning and pursuing meaningful family-related situations

albeit to a lower extent. The coefficients might underestimate the strengths of the convergent

relationships a bit due to a somewhat lowered reliability in this sample. While the rank order

of the Cronbach Alpha of the three OTH-scales was the same than in two samples used for

determining the factorial structure the actual coefficients were slightly lower (α = .72, α =

.59, and α = .74, for pleasure, engagement, and meaning, respectively).

Discussion

The present paper described the German adaptation of the Orientations to Happiness

measure of Peterson et al. (2005). The German OTH proved to be reliable and stable (within a

period of six month). Self- and peer-ratings converged well underlying the validity of the

scale. In the German and in the American version, four items had secondary loadings on other

factors than intended, especially item 9 (life of engagement). In the German OTH this item

loaded higher on P than on E4. Nevertheless, it was decided to maintain this item in order to

keep the German version identical with the US version. Nevertheless, this effect should be

taken into account when interpreting results (e.g., scale intercorrelations, factor structure).

The discrepancies do not seem to be a matter of translation but more a matter of different

cultural interpretations of the item. The fact that the CFA failed to yield a satisfying structure

raises the question whether a revision of the original OTH is indicated. After all, the German

factor structure is highly comparable to the US version (see below) but neither of the

4 A further translation of this item was tested using an Internet survey (N = 4,872, unpublished raw data) to be able to exclude translation problems as a reason for the double loading. The new translation underscored the “engagement character” of the item more directly. However, the loading pattern remained the same.

Orientations to happiness

16

theoretically well-justified models provided a good fit to the data. If such an endeavor is

undertaken it will be necessary to write items that load in the intended factor only.

Overall, the original factor structure found for the English language version of the OTH

was replicated. When comparing one to four factor models, the analyses suggested that a

three- and a four-factor solution yield good fit indices. Four factors might be applicable to

explain the data (e.g., by splitting the meaningful life into two dimensions—for example,

aspects relating to the society and to the self) but the authors favor the more parsimonious

three-factor solution. Furthermore, the factor solutions in the paper-pencil and in the Internet

samples were almost identical (Tucker’s Phi = .99 for all three orientations). This indicates

that the two samples yield equal factor structures, implying that the German OTH measure the

same constructs independently from the method of data collection used.

The scale intercorrelations were mainly similar to the English language version, yet

there were differences as well. As in previous studies involving the OTH (Peterson et al.,

2005; Peterson et al., 2007), the lowest correlation was found between P and M paper-pencil

and Internet data). Unexpectedly, the intercorrelations between P and E as well as M and E

showed coefficients comparable in size in the German-speaking samples. The intercorrelation

between P and E in the German form was affected by the double loading of one engagement-

item on E and P. When the analyses were done without item 9, the intercorrelation of E and P

decreased considerably.

As in the English version, younger and not married participants endorsed the

pleasurable life to a higher degree. Peterson et al. (2005) also found negative relations

between the pleasurable life and educational level, which was not replicated in the present

data. The latter might be due to different coding schemes for the educational level because of

the differences in the educational systems in the U.S. and the German speaking countries.

Orientations to happiness

17

Religiousness was introduced as a new variable. As expected, people practicing their religion

were lower in endorsement of the life of pleasure but higher in the life of meaning.

The correlations with satisfaction with life replicated findings with the English version

of the OTH. As expected, all three scales are related to subjective well-being supporting the

validity of the measure. Additionally, the convergent and discriminant validity of the OTH

was demonstrated by correlating the scales to both peer-reports and behavior in prototypical

situations. The latter showed that higher scores in one of the three orientations (i.e., pleasure,

engagement, and meaning) are related to self-reported typical behavior (self-reported time

spent on planning and pursuing and on wishing to ideally spend more time) in the prototypical

homologous situation while the correlations between heterologous variables were typically

non-significant (or at least much lower).

Further studies are planned in order to provide more information about the validity of

the German OTH. The relationship with satisfaction with life will be of further interest. As

the results obtained with the English form were replicated, the three orientations will be used

as a key to enhance “happiness” through interventions based on positive psychology

(Seligman, Steen, Park, & Peterson, 2005). Additionally, cross-cultural differences among the

German-speaking countries and other countries might provide information about different

cultural perspectives on the orientations to happiness (see Park, Peterson, & Ruch, 2009).

Furthermore, the intra-individual stability of the orientations to happiness needs to be

examined as well as the usefulness of the OTH in different areas of life (e.g., work, family,

leisure time).

18

References

Bollen, K. A. (1989). Structural equation models with latent variables. New York: John

Wiley.

Butcher, J. M., & Pancheri, P. (1976). A handbook of cross-national MMPI research.

Minneapolis, MN: University of Minnesota Press.

Coyne, I., & Bartram, D. (2006). ITC guidelines on computer-based and Internet-delivered

testing [Special Issue]. International Journal of Testing, 6(2).

Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York:

HarperCollins.

Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social

Indicators Research, 31, 103-157.

Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction with Life

Scale. Journal of Personality Assessment, 49, 71-75.

Gosling, S. D., Vazire, S., Srivastava, S., & John, O. P. (2004). Should we trust web-based

studies? A comparative analysis of six preconceptions about Internet questionnaires.

American Psychologist, 59, 93-104.

Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. (2006). Multivariate data analysis.

Upper Saddle River, NJ: Prentice-Hall.

Hancock, G. R. & Mueller, R. O. (2006). Structural equation modeling. Greenwich, CT:

Information Age Publishing.

Hu, L. T., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to

underparametrized model misspecification. Psychological Methods, 3, 424-453.

Muthén, L. K., & Muthén, B. (2007). Mplus User’s Guide (5th ed.). Los Angeles, CA: Muthén

& Muthén.

Orientations to happiness

19

Park, N., Peterson, C., & Ruch, W. (2009). Orientations to happiness and life satisfaction in

twenty-seven nations. The Journal of Positive Psychology, 4, 273-279.

Peterson, C., Park, N., & Seligman, M. E. P. (2005). Orientations to happiness and life

satisfaction: The full versus the empty life. Journal of Happiness Studies, 6, 25-41.

Peterson, C., Ruch, W., Beermann, U., Park, N., & Seligman, M. E. P. (2007). Strengths of

character, orientations to happiness, and life satisfaction. The Journal of Positive

Psychology, 2, 149-156.

Seligman, M. E. P. (2002). Authentic happiness. New York: Free Press.

Seligman, M. E. P., & Csikszentmihalyi, M. (2000). Positive psychology: An introduction.

American Psychologist, 55, 5-14.

Seligman, M. E. P., Steen, T. A., Park, N., & Peterson, C. (2005). Positive psychology

progress: Empirical validation of interventions. American Psychologist, 60, 410-421.

Tucker, L. R. (1951). A method for synthesis of factor analysis studies (Personnel Research

Section Report No. 984). Washington, DC: Department of Army.

20

Table 1

Fit Indices of the One- to Four-Factorial Solutions in the Paper-Pencil and Internet Sample

Model CMIN df p RMSEA SRMR

Paper-Pencil Sample

1-factor 2303.40 69 < .000 .168 .122

2-factor 1099.55 79 < .000 .106 .067

3-factor 543.39 75 < .000 .074 .043

4-factor 410.31 66 < .000 .067 .037

Internet Sample

1-factor 9534.86 75 < .000 .174 .131

2-factor 4019.85 88 < .000 .103 .065

3-factor 1706.80 86 < .000 .067 .038

4-factor 1054.22 76 < .000 .056 .031

Note. NPaper-pencil = 1,152. NInternet = 4,174.

RMSEA = Root Mean Square Error of Approximation. SRMR = Standard Root Mean Square

Residual.

Orientations to happiness

21

Table 2

Standardised Factor Loadings for the Orientations to Happiness Items in the Paper-Pencil

(PP) and in the Internet Sample (I) by Factor Analysis With VARIMAX Factor Rotation

Pleasure Engagement Meaning

Items PP I PP I PP I

Pleasure 3 .57 .58 .16 .15 .16 .12

8 .58 .57 .38 .38 .09 .13

13 .70 .69 .08 .06 .02 -.04

15 .60 .64 .12 .00 .02 -.06

16 .40 .36 .11 .11 .18 .26

18 .71 .71 -.01 -.02 -.09 -.07

Engagement 1 .16 .04 .33 .38 .11 .12

4 .16 .18 .39 .40 .33 .28

6 .07 .08 .68 .66 .06 .05

7 .16 .12 .76 .75 .11 .17

9 .37 .40 .35 .35 .13 .09

10 .07 .06 .36 .33 .08 .06

Meaning 2 .00 .02 .13 .11 .70 .78

5 .00 -.07 .29 .34 .40 .39

11 .04 .00 .11 .24 .63 .60

12 .06 .09 .17 .12 .72 .77

14 .05 .04 .13 .26 .66 .58

17 .12 .03 -.02 -.01 .36 .44

Note. Npaper-pencil = 1,152. Nonline = 4,174. Expected loadings are bold.

Orientations to happiness

22

Table 3

Descriptive Statistics and Intercorrelations of the Scales of the German OTH

Pleasure Engagement Meaning

PP I PP I PP I

Descriptives

M 3.20 3.27 2.95 3.18 2.95 3.16

SD .71 .70 .61 .62 .76 .79

Alpha .76 .73 .64 .63 .74 .75

rcmin .35 .29 .27 .27 .30 .32

rcmax .58 .54 .56 .56 .58 .59

Retest (3) .74 – .63 – .73 –

Retest (6) .77 – .63 – .70 –

Intercorrelations a

Engagement .42** .37**

Meaning .14** .12** .34** .35**

Age -.19** -.19** .05 .08** -.01 .02

Gender .08** -.01 -.01 -.05** .02 .04*

Married -.09** -.14** .05 .04** -.01 .02

Education -.06** -.03 .08** .08** .04 .10**

Religiousness -.11* -.13** -.02 .04* .16** .32**

Note. NPaper-pencil = 540-1,212. NInternet = 3,089-4,174. PP = Paper-pencil sample. I = Internet

sample. rcmin / rcmax = lowest/highest corrected item total correlation. Retest reliabilities were

for three (Retest 3; N = 113) and six month (Retest 6; N = 96). Gender (1 = male; 2 = female);

Married (0 = no; 1 = yes); education (1 = less than compulsory education to 5 = University

degree); Religiousness (practicing one’s religion: 2 = yes; 1 = no). a Pearson correlation; intercorrelations among the OTH scales are controlled for age, gender,

and education.

*p < .05. **p < .01.

Orientations to happiness

23

Table 4

Correlations Between Self- and Peer-Ratings in the OTH

Self-ratings Peer-ratings

Pleasure Engagement Meaning

Pleasure .51** .22* -.13

Engagement .19* .48** .26**

Meaning -.16 .05 .51**

Note. N = 125-127.

Orientations to happiness

24

Table 5

Correlations between the OTH and Measures for Indicators of the Experience of Pleasure in

Leisure-Situations, of Engagement in Work-Situations, and of Meaning in Family-Related

Situations

Pleasure Engagement Meaning

Leisure (Pleasure) Planning

.25**

.15*

.06

Pursuing .29** .10 .04

Ideal .26** .05 .07

mult R .34** .15 .08

Work (Engagement)

Planning

.12

.37**

.10

Pursuing -.03 .24** .07

Ideal .14 .22** .15*

mult R .21* .39** .16

Family (Meaning) Planning

.12

.21**

.23**

Pursuing .07 .19** .22**

Ideal .13 .14 .31**

mult R .14 .22* .40**

Note. N = 182-201. mult R = multiple correlation coefficient.

*p < .05; **p < .01.

Orientations to happiness

25

Appendix: Items of the German OTH

Scoring key: Pleasure = mean of items 3, 8, 13, 15, 16, and 18; Engagement = mean of items

1, 4, 6, 7, 9, and 10; Meaning = mean of items 2, 5, 11, 12, 14, and 17.

A copy of the scale can be obtained by writing to the first author.

Scale Item number Item

3 Das Leben ist zu kurz, um die Freuden, die es bietet, aufzuschieben. 8 Ich tue alles dafür, um mich in Hochstimmung zu versetzen.

13 Bei der Auswahl von Aktivitäten ist es mir immer wichtig, dass sie Spaß machen.

15 Ich stimme der Aussage zu: „Das Leben ist kurz – iss das Dessert zuerst“.

16 Ich tue gerne Dinge, die meine Sinne anregen. 18 Für mich bedeutet ein gutes Leben ein Leben, das Vergnügen bereitet.

Pleasure

1 Was ich auch tue, die Zeit vergeht wie im Flug. 4 Ich suche mir Situationen, die meine Fähigkeiten und mein Geschick

herausfordern 6 Ob bei der Arbeit oder beim Spiel, ich bin mit Leib und Seele dabei

und mir meiner selbst nicht bewusst. 7 Ich gehe immer völlig in dem auf, was ich tue. 9 Bei der Auswahl von Aktivitäten berücksichtige ich immer, ob ich

mich in ihnen verlieren kann. 10 Ich lasse mich selten durch Dinge ablenken, die um mich herum

passieren.

Engagement

2 Mein Leben dient einem höheren Zweck. 5 Bei der Auswahl von Aktivitäten achte ich immer darauf, dass sie

anderen Menschen nützen. 11 Ich trage Verantwortung, die Welt besser zu machen. 12 Mein Leben hat einen bleibenden Sinn 14 Was ich tue, spielt eine Rolle für die Gesellschaft. 17 Ich denke gerne darüber nach, worin der Sinn des Lebens besteht und

wo mein Platz darin ist.

Meaning