supp.apa.org · Web view2021. 4. 30. · Let’s go forward with this new line. ... .039. 6. 2 cells. Difference of means. t (277) = 2.68, p = .008. ... Let’s make as much profit

1

Supplemental Online Materials for

Instrumental Use Erodes Sacred Values

Table of Contents: 1. Study 1 Materials and Supplemental Analyses (p. 2) 2. Study 2 Materials and Supplemental Analyses (p. 3)3. Supplemental Scale Validation Information (p. 6) 4. Study 3 Materials and Supplemental Analyses (p. 8) 5. Study 4 Supplemental Analyses (p. 17) 6. Study 5 Materials and Supplemental Analyses (p. 18)7. Study 6 Materials and Supplemental Analyses (p. 22) 8. Study 7 Materials and Supplemental Analyses (p. 26) 9. Results from the Purity Subscale in Studies 2, 4, 5, and 6 (p. 33) 10. Item-by-Item Results for Studies 1-6 (p. 35) 11. Summary of Studies not in the Paper (p. 40) 12. Power Analyses (p. 54)

2

Study 1

Materials:

[Control]The Flag of the United States consists of thirteen equal horizontal stripes of red (top and bottom) alternating with white, with a blue rectangle in the canton (referred to specifically as the "union") bearing fifty small, white, five-pointed stars. Nicknames for the flag include the "Stars and Stripes," "Old Glory," and "The Star-Spangled Banner."

[Instrumental Use Condition]Recently, an organization launched a "national pride" campaign. This campaign uses the national flag and anthem as a part of the company's advertising.

Although publicly said to reflect a love for civic spirit, the community, and national pride, the owners and operators of the car company conducted a series of market analyses that revealed that advertising using national symbols would be most profitable approach for the company.

[Value-consistent Use Condition]Recently, a community organization launched a "national pride" campaign. This campaign uses the national flag and anthem as a part of the company's advertising.

This campaign reflects the organization's long-standing desire to facilitate civic spirit, the community, and national pride.

Supplemental Analyses:

In both Study 1 in the main text (the replication study), and Study 1 in this document (the

original study), we note that the campaign in the value-consistent use condition was still seen as

somewhat motivated by profit (M of 4.08 out of 7). There was also more variance in this

condition (SD = 1.40) than in the instrumental use condition (SD = 1.02). There are likely

important individual differences that drive how people perceive even seemingly value-consistent

campaigns, such as general cynicism or distrust in organizations. Importantly, when we look at

the correlations between the manipulation check and sacredness, there is a significant, negative

correlation between perceive the campaign as motivated by profit and sacredness, r = -.21, p

= .003, which supports our central prediction.

3

Study 2

Experimental Materials:

Please see https://osf.io/mfpzx/?view_only=4df1d286d00443bd92431cb4d7cfbe00 for tweet images.

Please indicate how well you think each of the following describes your beliefs about the issue of environmentalism. I believe that environmentalism is…

Strongly Disagree

Disagree Somewhat Disagree

Neither Agree nor Disagree

Somewhat Agree

Agree Strongly Agree

Pure o o o o o o oTainted o o o o o o oClean o o o o o o o

Sacred o o o o o o oShould not be compromised, no matter the benefits (money or otherwise)

o o o o o o o

I revere, respect it

o o o o o o o

Involves principles that I would defend under any circumstances

o o o o o o o

Cannot be contaminated by financial or other commercial interests

o o o o o o o

https://osf.io/mfpzx/?view_only=4df1d286d00443bd92431cb4d7cfbe00

4

Involves issues or values I could compromise if the situation demands it

o o o o o o o


In analyzing the control variables, only political orientation significantly predicted

sacredness, b = -.10, p = .026, with more conservative participants holding environmental

protection as less sacred. This finding is consistent with past research (e.g., Dunlap, Xiao, &

McCright, 2001; Feinberg & Willer, 2013). Similarly, engagement with Earth Day activities

predicted sacredness, b = .38, p < .001, with more engagement predicting greater sacredness.

Because neither variable significantly interacted with the predictor variable (ps > .209), nor did

controlling for these variables affect the results, these variables are not discussed further.

In our revised version of Study 2, we included two exploratory items to assess the

potential attributes of these campaigns that may be shaping instrumentality perceptions. We

included two items assessing the organization’s for-profit orientation (“In general, the

organization that sent this message is more aligned with a profit motivation”) and perceived

congruence with environmentalism (“In general, the organization that sent this message is well-

aligned with the issue of environmentalism”) on a scale from 1 (Strongly disagree) to 7

(Strongly agree). Perceptions of instrumentality (“The organization sent this message because

they thought it would be profitable to do so”) were significantly related to both perceptions of a

for-profit orientation, r = .72, p < .001, and misalignment with environmentalism, r = -.37, p

< .001. Though the central concern in the current work was perceptions of instrumentality itself,

5

and the primary goal in Study 2 was to examine how people encounter instrumental use in the

real world, further work unpacking the different attributes that shape perceptions of

instrumentality will be important going forward.

6

Supplemental Scale Validation Information

Given that we made some adjustments to an existing scale (Hanselmann & Tanner,

2008), as well as adjustments during this project, we offer additional scale information below.

Moreover, as raised by one reviewer, it is possible that participants are interpreting the three

purity items (“pure,” “tainted,” and “clean”) as indicating that a sacred value has experienced a

violation rather than a loss in sacredness. This helpful suggestion prompted additional

exploration into the factor structure underlying our scales. Because purity appeared to capture a

distinct factor empirically, and because of the aforementioned conceptual reasons, the results in

the main text analyze uncompromisability factor only. The factor loadings for the second version

of the scale are reported in text in the Appendix.

Additional Exploratory Factor Analyses:

Study 3 (Scale from Study 2 used). Conducting an exploratory factor analysis with direct

oblimin rotation revealed a two-factor solution. The eigenvalues for the first factor was 4.35

(factor loadings > .17), and 1.14 for the second factor (factor loadings > .21). The purity and

uncompromisability subscales showed good internal consistency (s = .88 and .76, respectively).









7





Confirmatory Factor Analysis:

We also conducted a confirmatory factor analysis to test the two-dimensional structure of

the scale in Studies 2-6 using the Lavaan package in R, again using the study with the largest

sample (Study 5). Overall goodness-of-fit statistics suggest that our data fit the model well

(χ2(689) = 116.90, p < .001; root mean square error of approximation (RMSEA) = .07;

confirmatory fit index (CFI) = .97; and standardized root-mean-square residual (SRMR) = .05).

Values of up to .08 for RMSEA and SRMR and greater than .90 for CFI are considered

acceptable (Bagozzi & Yi, 1988; Hu & Bentler, 1999). To further assess the appropriateness of

the two-factor model, we compared the goodness of fit of the two-factor model to a single-factor

model. The two-factor model demonstrated a significant improvement in overall fit over the

single-factor model Δ χ2 = 859.67, p < .001.

8

Study 3

Study Materials:[Instrumental use]

The CEO of a company spoke to the company's board of directors and said, "We are thinking about launching a new pro-environmental program that will sell us to the world as an eco-friendly organization."

The chairman of the board then replied, "Good, this 'green' stuff sells. Let's make as much profit as we can off of this."

Another board member then cited a recent report asking similar organizations who had launched pro-environmental campaigns why they had done so. The organizational members who participants in the survey could choose which of the following issues MOST motivated their company's environmental campaigns-

The response options were: A) Reduced Waste – Sustainability programs help cut down on waste in the organization B) Sustainability – It is critically important to help the environment C) Profit – It helps the bottom line and facilitates profit D) Improved Workforce – Sustainable workplaces are better for employees, OR E) Public Relations – It will attract like-minded consumers

[High Prevalence Condition]

The results suggested that 90% of similar organizations launched their environmental campaigns in order to yield a profit.

[Low Prevalence Condition]

The results suggested that 10% of similar organizations launched their environmental campaigns in order to yield a profit. The majority of organizations launched their campaigns for other reasons.

[Value-consistent use condition]

The CEO of a company spoke to the company's board of directors and said, "We are thinking about launching a pro-environmental campaign that will promote the importance of an eco-friendly lifestyle."

The chairman of the board then replied, "Good, let's increase our commitment to environmental cause as much as we can."

9

Another board member then cited a recent report asking members of similar organizations who had launched pro-environmental campaigns to report why they had done so. The organizations that completed the survey could choose which of the following issues MOST motivated their organization's environmental campaigns-

The response options were: A) Reduced Waste – Sustainability programs help cut down on waste in the organization B) Sustainability – It is critically important to help the environment C) Profit – It helps the bottom line and facilitates profits D) Improved Workforce – Sustainable workplaces are better for employees, OR E) Public Relations – It will attract like-minded consumers

[High Prevalence Condition]

The results suggested that 90% of similar organizations launched their environmental campaigns with the primary aim of increasing sustainability, i.e., helping the environment. In other words, the majority of organizations launched their campaigns to support the environment and sustainability.

[Low Prevalence Condition]

The results suggested that 10% of similar organizations launched their environmental campaigns with the primary aim of increasing sustainability, i.e., helping the environment. The majority launched their campaigns for other reasons.

Please indicate how well you think each of the following describes your beliefs about the issue of environmentalism.

Strongly Disagree

Disagree Somewhat Disagree

Neither Agree nor Disagree

Somewhat Agree

Agree Strongly Agree

Pure o o o o o o oTainted o o o o o o oClean o o o o o o o

Sacred o o o o o o oShould not be compromised

o o o o o o o

10

, no matter the benefits (money or otherwise)

I revere, respect it

o o o o o o o

Involves principles that I would defend under any circumstances

o o o o o o o

Cannot be contaminated by financial or other commercial interests

o o o o o o o

Involves issues or values I could compromise if the situation demands it

o o o o o o o

Thinking about the situation you read about in Part 1, how do you feel right now, at this point in time?

Not at all 1

2 3 Somewhat

4

5 6 Very much so 7

Disgusted o o o o o o oUpset o o o o o o oSaddened o o o o o o oOutraged o o o o o o oAngry o o o o o o o

11

Offended o o o o o o oPleased o o o o o o oUncomfortable o o o o o o oUneasy o o o o o o oBothered o o o o o o oIrritated o o o o o o oAnnoyed o o o o o o o

Thinking again about the situation you read about in Part 1, please rate the leader's actions along the following dimensions:

Not at all

1

2 3 Somewhat

4

5 6 Very much so 7

How wrong was this act?

o o o o o o o

How severe was this act?

o o o o o o o

How harmful [i.e., involving physical and/or emotional suffering] was this act?

o o o o o o o

How impure [i.e., involving sinfulness, indecency, dirtiness]

o o o o o o o

12

was this act?

How certain are you of your rating of the act's harmfulness?

oNot at all certain 1

o 2

o 3

o Somewhat certain 4

o 5

o 6

oVery certain 7

How certain are you of your rating of the act's impurity?

oNot at all certain 1

o 2

o 3

o Somewhat certain 4

o 5

o 6

oVery certain 7

Please complete the following with regard to the CEO and board of directors' communication that you just read about.

The organization is trying to make my decisions for me as a consumer.

oNot at all 1

13

o 2

o 3

o Somewhat 4

o 5

o 6

oVery much so 7

The organization is trying to manipulate me as a consumer.

oNot at all 1

o 2

o 3

o Somewhat 4

o 5

o 6

oVery much so 7

The organization is trying to pressure me as a consumer.

oNot at all 1

o 2

o 3

o Somewhat 4

o 5

o 6

14

oVery much so 7

Please recall the message you viewed in Part 1 of this study. What was main motivation behind the organization's decision to launch the pro-environmental campaign?

o To make money

o To encourage environmental protection

o It did not say

Donation Measure:

Thank you for your participation.

As a part of doing behavioral research relevant to judgments and values, a large mission for our lab is to make donations available to various nonprofits. In most of our studies, we like to give people the option to participate in these charitable causes. As such, you have been provided with an additional $0.25 (on top of your base pay), which you may now choose to contribute to charity, or will receive after the study is complete. You can donate as much or as little of the money as you want ($0-0.25).

If you choose to donate, you will be provided with a list of charities to choose from on the next page.

How much would you like to donate today? ______

If you chose to donate:

Thank you for your support. Please select which of the following you choose to donate to: ____ The Sierra Club: A non-profit geared toward the preservation of land and forest, clean air and water, and a host of other issues (see www.sierraclub.org).____ Get Fit Foundation: A non-profit geared toward promoting active, healthy lifestyles (see more at getamericafit.org).____ Teaching Tolerance: A non-profit dedicated to promoting diversity and creating equitable school experiences for our nation's children (see more at tolerance.org).


As indicated in the materials above, we also included exploratory measures of affect

using 12 items on 7-point scales and reactance using 3 items (e.g., “This organization is trying to

manipulate me as a consumer”) on 7-point scales. Moreover, we collected data assessing the

15

perceived harmfulness (“how harmful [i.e., involving physical and/or emotional suffering] was

this act?”) and impurity (“how impure [i.e., involving sinfulness, indecency, dirtiness] was this

act?”) on a scale from 1 (Not at all) to 7 (Very much so).

Negative Affect. A two-way ANOVA revealed no interaction between motive

(instrumental use vs. value-consistent use) and prevalence (high vs. low) conditions in predicting

negative affect, F(1, 390) = 2.65, p = .104, ɳ2 = .007, a significant effect of motive condition,

F(1, 390) = 87.46, p < .001, ɳ2 = .183, such that participants experienced more negative affect in

the instrumental use condition, and no effect of prevalence condition, F(1, 390) = .42, p = .517,

ɳ2 = .001.

Dissonance. Included in the negative affect items were three items (uncomfortable,

uneasy, bothered), which have been used in prior research to capture feelings of dissonance (e.g.,

Glasford, Pratto, & Dovidio, 2008). A two-way ANOVA revealed no interaction between motive


negative affect, F(1, 390) = 1.46, p = .228, ɳ2 = .004, a significant effect of motive condition,

F(1, 390) = 66.25, p < .001, ɳ2 = .145, such that participants experienced more dissonance in the

instrumental use condition, and no effect of prevalence condition, F(1, 390) = 2.23, p = .136, ɳ2 =

.006.

Reactance. A two-way ANOVA revealed a marginal interaction between motive


reactance, F(1, 389) = 3.01, p = .083, ɳ2 = .008, a significant effect of motive condition, F(1,

389) = 115.69, p < .001, ɳ2 = .229, such that participants experienced more reactance in the

instrumental use condition, and no effect of prevalence condition, F(1, 389) = .001 p = .975, ɳ2

= .000.

16

Harm vs. Purity Violation. In terms of perceiving a harm violation, a two-way ANOVA

revealed no interaction between motive (instrumental use vs. value-consistent use) and

prevalence (high vs. low) conditions in predicting harm perception, F(1, 389) = 0.08, p = .779, ɳ2

= .000, a significant effect of motive condition, F(1, 389) = 36.91, p < .001, ɳ2 = .087, such that

more harm was perceived following instrumental use, and no effect of prevalence condition, F(1,

389) = .04 p = .842, ɳ2 = .000. In terms of perceiving a purity violation, a two-way ANOVA

revealed no interaction between motive (instrumental use vs. value-consistent use) and

prevalence (high vs. low) conditions, F(1, 389) = 1.08, p = .229, ɳ2 = .003, a significant effect of

motive condition, F(1, 389) = 164.73, p < .001, ɳ2 = .297, such that participants were more likely

to perceive a purity violation following instrumental use, and no effect of prevalence condition,

F(1, 389) = .001 p = .976, ɳ2 = .000.

Mediation. We note that, while the effect size for donations was small, we tested a

mediated moderation model using Model 7 in PROCESS, and found that there was a significant

indirect effect of the motive x prevalence interaction on donation through a decrease in perceived

sacredness, 95% CI [-.018, -.001].

Testing a Dissonance Account. Given the boost in both reactance and feelings of

dissonance following instrumental use, we testing whether these variables may drive the

decrease in sacredness, and offer an alternative mechanism for our findings. We tested the

potential role of both feelings of dissonance and reactance again using Model 7 in PROCESS,

and found that there was no significant indirect effect of the motive x prevalence interaction on

sacredness through feelings of dissonance, 95% CI [-.018, -.001], or psychological reactance,

95% CI [-.257, .007].

17

Study 4


Consistent with Tetlock et al.’s (2000) findings, condition significantly interacted with

political ideology to predict donation to the diversity cause. Specifically, regressing whether or

not participants donated to the diversity cause onto the taboo tradeoff condition (instrumental use

= 0; taboo tradeoff = 1; value-consistent use = 0) and participants’ self-reported social

conservatism-liberalism revealed no effect of condition, b = -0.05, p = .114, a marginal effect of

political orientation, b = -0.008, p = .075, and a significant interaction, b = 0.20, p = .008. We

used a floodlight analysis to determine the range(s) of political orientation for which the effect of

the manipulation was significant (Spiller, Fitzsimons, & Lynch, 2013). The results revealed that

participants were significantly more likely to donate to the diversity cause in the taboo tradeoff

condition than in the other two conditions if they scored 3.70 (out of 7) or higher on the scale, bjn

= 0.03, SE = 0.01, z = 2.20, p = .029, with higher scores representing a more liberal orientation.

Thus, just as Tetlock et al. found that only liberals reacted with outrage to diversity-relevant

taboo tradeoffs, we found that only liberals were more likely to donate after exposure to a taboo

tradeoff. However, consistent with Studies 1 through 3 in the current paper, we again found no

such interaction between the instrumental use condition and political orientation, b = -0.003, p

= .731, suggesting participants were similarly affected by instrumental use regardless of political

orientation.

18

Study 5

Study Materials:

Part 1: Organizational Decision-Making

Thank you for your participation. You will now read about an exchange that occurred in an organization. Please read carefully as you will be asked to recall this information later.

[Value-consistent use condition]

The CEO of a company spoke to the company’s board of directors about tactics to boost the company’s sustainability (i.e., how they could increase their positive environmental impact). One idea the CEO proposed was to introduce a new “green” or environmentally friendly campaign.

One specific idea the CEO proposed was to include a new line of green or sustainable products. The company’s analysts estimated that the new product line would substantially decrease water usage. These products would decrease waste and environmental harm (i.e., would improve sustainability). The chairman of the board replied, “Good, being green matters. Let’s go forward with this new line.”

[Instrumental support condition]

The CEO of a company spoke to the company’s board of directors about tactics to boost the company’s bottom line (i.e., increase profits). One idea the CEO proposed was to introduce a new “green” or environmentally friendly campaign.

The CEO proposed to include a new line of "green" or sustainable products. The company’s analysts estimated that the new product line would increase net earnings. These products would decrease waste and environmental harm (i.e., would improve sustainability), while also creating value for shareholders and the company. The chairman of the board replied, “Good, being green is profitable right now. Let’s go forward with this new line.”

[Instrumental threat condition]

The CEO of a company spoke to the company’s board of directors about tactics to boost the company’s bottom line (i.e., increase profits). One idea the CEO proposed was to introduce a new “green” or environmentally friendly campaign.

The CEO proposed to include a new line of "green" or sustainable products. The company’s analysts estimated that the new product line would increase net earnings. Being concerned about profitability, the new line of products still used cheaper, plastic materials. These products would

19

increase waste and environmental harm, but would create value for shareholders and the company. The chairman of the board replied, “Good, being green is profitable right now. Let’s still sell ourselves as green anyway.”

[Taboo Tradeoffs]

The CEO of a company spoke to the company’s board of directors about a decision. The CEO and board of directors just learned that the company was using plastic products that were polluting the environment and creating a large amount of waste and environmental harm.

The CEO can stop the pollution, but it would cost the company - money which could be spent in other ways, such as purchasing equipment, increasing marketing, and increasing the firm's technological advantage.

The chairman of the board replied that they should continue creating waste and environmental harm, to instead keep the money for other organizational purposes.

To what extent does this organization's actions uphold the value of environmentalism?

oNot at all 1

o 2

o 3

o 4

o 5

o 6

oVery much so 7

To what extent does this organization's actions violate the value of environmentalism?

oNot at all 1

o 2

o 3

o 4

20

o 5

o 6

oVery much so 7

To what extent should people be concerned about what this action means for the value of environmentalism?

oNot at all 1

o 2

o 3

o 4

o 5

o 6

oVery much so 7

Thinking again about the situation you read about in Part 1, please rate the organization's actions along the following dimensions:

Not at all

1

2 3 Somewhat

4

5 6 Very much so 7

How wrong was this act?

o o o o o o o

How severe was this act?

o o o o o o o

How harmful [i.e., involving physical and/or emotional

o o o o o o o

21

suffering] was this act? How impure [i.e., involving sinfulness, indecency, dirtiness] was this act?

o o o o o o o


Moderation. To assess whether the effect of condition was moderated by political

orientation, we conducted a moderated regression, regressing participants’ sacredness ratings

onto three dummy-coded condition variables, their political orientation, and the interactions

between the condition variables and political orientation. Only the effect of the instrumental

threat condition significantly interacted with political orientation to predict sacredness, binteraction =

0.20, t = 2.80, p = .005, bpolitical = -.22, t = -4.29, p < .001, bthreatcondition = -0.64, t = -2.24, p = .025.

Following the recommendations of Spiller et al. (2013), we again used the Johnson-Neyman

technique to identify the range(s) of political orientation for which the simple effect of the

manipulation (1 = instrumental threat condition; 0 = the other conditions). The analysis revealed

that participants held the value as significantly more sacred in the instrumental threat condition

than in the other conditions if their political orientation score was 4.41 or higher (bjn = .22, SE

= .11, p = .05). Thus, the instrumental threat condition bolstered sacredness among more

conservative participants.

Harm vs. Purity Violation. In terms of perceiving a harm violation, there was a

significant effect of condition on perceived harm, such that more harm was perceived in the

taboo tradeoff condition, followed by the instrumental threat condition, then the instrumental

22

support condition, and the value-consistent use condition. All conditions significantly differed

from each other (see Table S1). Examining perceptions of a purity violation, participants

perceived more of a purity violation in the taboo tradeoff condition and the instrumental threat

conditions (these two conditions did not significantly differ from each other), followed by the

instrumental support, and value-consistent use conditions.

Table S1. Means, Standard Deviations, and Statistical Tests for Perceptions of Harm and Purity Violations (Study 5).

Study 6

Study Materials:

Imagine now that [the CEO of the hospital system] is confronted with a resource allocation decision. The children's wing of the hospital is not profitable. Keeping the wing open will cost the hospital millions of dollars that could be spent in other ways, such as purchasing better equipment and enhancing salaries to recruit talented doctors to the hospital. However, closing the wing would mean that many sick kids would have to be moved and would be temporary without a place for treatment. [The CEO] has decided to close the children's wing to save money for other organizational purposes.

According to how you feel right now, at this point in time, to what degree to each of the following words describe how you would feel toward [the CEO] if he decided to close the children's wing to save money for other organizational purposes?

Not at all 1

2 3 Somewhat 4

5 6 Very much so 7

Disgusted o o o o o o oUpset o o o o o o o

23

Outraged o o o o o o oAngry o o o o o o oNeutral o o o o o o oUncomfortable o o o o o o oUneasy o o o o o o oBothered o o o o o o o

Please indicate the degree to which you agree with the following statements.

[The CEO] should be removed from his job.

o Strongly agree

oAgree

o Somewhat agree

oNeither agree nor disagree

o Somewhat disagree

oDisagree

o Strongly disagree

[The CEO] should not be punished for his decision.

o Strongly agree

oAgree

o Somewhat agree

24


o Somewhat disagree

oDisagree

o Strongly disagree

If [the CEO] was a friend of mine, and I knew the decision he made, I would end the friendship over this issue.

o Strongly agree

oAgree

o Somewhat agree


o Somewhat disagree

oDisagree

o Strongly disagree

25

Thinking back now to the video you watched in Part 1 of the study, how did you feel?Not at all 1

2 3 Somewhat 4

5 6 Very much so 7

Disgusted o o o o o o oUpset o o o o o o oSaddened o o o o o o oOutraged o o o o o o oAngry o o o o o o oOffended o o o o o o oPleased o o o o o o oUncomfortable o o o o o o oUneasy o o o o o o oBothered o o o o o o oInspired o o o o o o oGrateful o o o o o o o

Supplemental Results:

Positive and negative affect. In examining emotional reactions to the video, participants

in the value-consistent use condition experienced more positive affect (M = 4.55, SD = 1.72)

while watching the video than did those in the instrumental use condition (M = 4.18, SD = 1.67),

t(275) = 1.81, p = .072, though this difference was not significant. There were no significant

differences in experience negative affect between the value-consistent use (M = 2.32, SD = 1.51)

and instrumental use (M = 2.48, SD = 1.46), t(275) = 0.90, p = .367. This is consistent with our

26

theorizing that instrumental use (especially a subtle use such as a corporate sponsorship) does not

elicit a sufficient degree of negative affect to lead to value protection effects.

Dissonance. Finally, we again also examined the results of the dissonance items

separately and found no difference between the value-consistent use (M = 2.42, SD = 1.69) and

instrumental use (M = 2.67, SD = 1.69) conditions, t(275) = 1.22, p = .223.

Study 7

Study Materials:

To what degree are you a fan of the National Football League (NFL)?

oNot at all 1

o 2

o 3

o Somewhat 4

o 5

o 6

oVery much so 7

To what degree are you a fan of football, more generally?

oNot at all 1

o 2

o 3

o Somewhat 4

o 5

27

o 6

oVery much so 7

About how many televised NFL games would you watch in a given football season?

________________________________________________________________

About how many live (in-person) NFL games would you watch in a given football season?

________________________________________________________________

Do you have a favorite NFL team?

oYes

oNo

Which is your favorite team?

________________________________________________________________

FOOTBALL TRIVIA CHALLENGE! In this first section, you will be asked a series of trivia questions in TRUE or FALSE format. You will be provided with the answers at the end of the study.

Note that you will have a maximum of 45 seconds to answer each question before the page moves forward.

[Critical Item]

Is it TRUE or FALSE that the term 'Paid Patriotism' was used to describe the practice of NFL teams accepting money from the military and other organizations to put on patriotic displays at games?

o True

o False

28

How confident are you in your answer on the previous page?

oNot at all confident 1

o 2

o 3

o Somewhat confident 4

o 5

o 6

oVery confident 7

How many questions do you think you got right (out of 5) on the NFL trivia?

o 1/5

o 2/5

o 3/5

o 4/5

o 5/5

How difficult did you find the trivia items?

oNot at all difficult 1

o 2

o 3

o 4

oVery difficult 5

29

POTENTIAL CHANGE IN NFL POLICYWe are collecting consumer data about responses to a potential change in NFL policy. The NFL is currently considering letting the decision to sing the national anthem be voluntary and up to individual teams rather than be mandatory across the NFL. What do you think about this proposal?

o I think this is a good thing.

o I do not object to this.

o I think this should be done only if it brings great benefits (e.g., a better game day experience).

oNo matter how great the benefits, this should not be done.

Would you be willing to write a letter of support (or a letter of protest) to the NFL about the previously described policy change?

oYes - I would be willing to write to the NFL

oNo - I would not be willing to write to the NFL

Which factors do you think influenced your responding to the potential policy change?

________________________________________________________________

________________________________________________________________

________________________________________________________________

________________________________________________________________

________________________________________________________________

30

Please now think of patriotic displays put on by National Football League teams (e.g., the anthem sung at the beginning of the game, celebration of the military).

To what extent do you think the NFL displays patriotism because they receive money to do so?

oNot at all 1

o 2

o 3

o Somewhat 4

o 5

o 6

oVery much so 7

To what extent do you think the NFL displays patriotism because they are truly patriotic?

oNot at all 1

o 2

o 3

o Somewhat 4

o 5

o 6

oVery much so 7

Additional demographics not included in all other studies:

Have you ever served in the US military?

oYes

31

oNo

In which country were you born?

________________________________________________________________

In which country do you currently reside?

________________________________________________________________


As indicated in our preregistration document, we also included an exploratory measure of

participants’ willingness to write a letter to the NFL (either in support or in protest of the

potential change). The results revealed no significant effect, χ2 (1, N = 484) = 2.79, p = .094, OR

= 1.50. 16.6% of participants who were aware of paid patriotism were willing to write a letter,

and 23.0% of participants who were not aware of paid patriotism were willing to write a letter.

We also conducted exploratory analyses to see whether awareness of paid patriotism

might interact with political orientation, NFL fandom, and confidence in their awareness of paid

patriotism (see Table 2 for correlations among the variables of interest). First, we conducted a

binary logistic regression in which we regressed participants’ tradeoff decision (1 = sacred value;

0 = nonsacred value) onto the dummy coded awareness variable (1 = aware; 0 unaware),

centered political orientation variable, and their interaction. The results revealed the predicted

effect of awareness of paid patriotism, b = -0.52, χ2 (1, N = 485) = 5.06, p = .025, a significant

effect of political orientation, b = -0.52, χ2 (1, N = 485) = 28.68, p < .001, and no significant

32

interaction, b = -.08, χ2 (1, N = 485) = 0.28, p = .594. Next, looking at fandom, the results

revealed the predicted effect of awareness of paid patriotism, b = -0.62, χ2 (1, N = 503) = 9.71, p

= .002, no effect of fandom, b = 0.10, χ2 (1, N = 503) = 1.53, p = .201, and no significant

interaction, b = -0.13, χ2 (1, N = 503) = 1.60, p = .207. Finally, examining confidence in their

answer for the paid patriotism question, the results revealed the predicted effect of awareness of

paid patriotism, b = -0.74, χ2 (1, N = 503) = 12.14, p < .001, a significant effect of confidence, b

= 0.24, χ2 (1, N = 503) = 5.95, p = .015, and a significant awareness x confidence interaction, b =

-0.39, χ2 (1, N = 503) = 10.74, p = .001. A floodlight analysis revealed that awareness was

related to reduced tradeoff resistance among participants with centered confidence scores greater

than -0.94. Thus, the effect emerged only among those more confident in their answers.

Table S2. Correlations among the variables of interest in Study 7.

33

Results from the Purity Subscale

Study 2

Purity. There was also a main effect of condition on purity judgments, F(2, 540) = 16.52,

p < .001, 2 = .058. Participants in the instrumental use condition rated environmentalism as

significantly less pure (M = 4.95, SD = 1.28) than did participants in the value-consistent use

condition (M = 5.60, SD = 0.93), t(350) = -5.18, p < .001, d = 0.55, and in the baseline control

condition (M = 5.44, SD = 1.06), t(364) = -4.21, p < .001, d = 0.44. Participants in the baseline

control condition rated environmentalism as similarly pure as those in the value-consistent use

condition, t(366) = 1.56, p = .120, d = 0.16.

Study 3

Purity. A two-way ANOVA revealed a significant interaction between motive


the perceived purity of environmentalism, F(1, 390) = 9.31, p = .002, ɳ2 = .023, a significant

main effect of motive condition, F(1, 390) = 37.06, p < .001, ɳ2 = .087, and a main effect of

prevalence condition, F(1, 390) = 4.50, p = .035, ɳ2 = .011. Parsing apart this interaction, in the

high prevalence condition, participants who read about instrumental use rated environmentalism

as significantly less sacred (M = 3.95, SD = 1.46) than did those in the value-consistent use

condition (M = 5.14, SD = 1.09), t(201) = -6.48, p < .001, d = 0.91. In the low prevalence

condition, participants who read about instrumental use rated environmentalism as marginally

less sacred (M = 4.63, SD = 1.28) than did those in the value-consistent use condition (M = 5.02,

SD = 1.24), t(189) = -2.14, p = .033, d = 0.31.

34

Study 4


p < .001, 2 = .108. Participants in the instrumental use condition rated diversity as significantly

less pure (M = 4.29, SD = 1.48) than did participants in the value-consistent use condition (M =

5.44, SD = 1.32), t(105) = -4.07, p < .001, d = 0.79, but did not differ from the taboo tradeoff

condition (M = 4.53, SD = 1.59), t(99) = -0.82, p = .411, d = 0.16. Participants in the taboo

tradeoff condition rated diversity as less pure than did participants in the value-consistent use

condition, t(94) = -3.04, p = .003, d = 0.66.

Study 5


p < .001, 2 = .030. Participants in the instrumental support condition rated environmentalism as

significantly less pure (M = 4.67, SD = 1.36) than did participants in the value-consistent use

condition (M = 5.22, SD = 1.19), t(342) = -3.70, p = .001, d = 0.40. Participants in the

instrumental support condition did not significantly differ from participants in the instrumental

threat condition (M = 4.62, SD = 1.42), t(336) = -0.31, p = .760, d = 0.03, or taboo tradeoffs

condition (M = 4.78, SD = 1.54), t(320) = -0.71, p = .480, d = 0.08, in purity judgments. The

value-consistent use condition differed from all other conditions (ps < .004), and no other

significant differences emerged.

Study 6

Purity. Participants in the instrumental use condition rated The Wave as significantly

less pure (M = 4.96, SD = 1.24) than did participants in the value-consistent use condition (M =

5.56, SD = 1.09), t(277) = -4.27, p < .001, d = 0.51.

35

Item-by-Item Results Across Studies

Table S3. Results of key statistical tests at the item level.

ItemRelevant studies

Study design

Key statistical

result Results

I believe that X is sacred 1 3 cellsDifference of means

F(2, 265) = 3.40, p = .035

2 3 cellsDifference of means F(2, 540) = 10.45, p < .001

3 2x2

2-way interaction, attenuation F(1, 390) = 1.17, p = .279


5 4 cellsDifference of means F(3, 659) = 5.23, p = .001

6 2 cellsDifference of means t(276) = 0.69, p = .493

I believe that X should not be sacrificed, no matter the benefits. 1 3 cells

Difference of means F(2, 265) = 3.84, p = .023


3 2x2

Two-way interaction (attenuation) F(1, 390) = 0.71, p = .399




X involves principles that I would defend under any circumstance. 1 3 cells



3 2x2



36



I feel committed to X 1 3 cellsDifference of means F(2, 265) = 5.09, p = .007

X is inviolable 1 3 cellsDifference of means

F(2, 265) = 2.73, p = .067

I revere, respect X 2 3 cellsDifference of means F(2, 540) = 6.46, p = .002

3 2x2





Cannot be contaminated by financial or other commercial interests 2 3 cells


3 2x2





Involves issues or values I could compromise if the situation demands it (Reverse-coded)

2 3 cells Difference of means

F(2, 540) = 6.02, p = .003

3 2x2 Two-way interaction

(attenuation)

F(1, 390) = 6.07, p =.014


F(2, 149) = 0.44, p=.648


F(3, 659) = 3.62, p = .013


t(277) = 1.26, p = .209

37

Table S4. Means and Standard Deviations in Study 1.

Instrumental Use

Value Consistent Control

Item M (SD) M (SD) M (SD)I believe that the flag and other national symbols are sacred. 3.83 (1.91) 4.49 (1.65) 4.42 (1.94)I believe that the flag and other national symbols should not be sacrificed, no matter the benefits. 3.68 (1.94) 4.46 (1.68) 4.23 (1.98)The flag and other national symbols involve principles that I would defend under any circumstance. 3.64 (1.72) 3.89 (1.59) 4.36 (1.61)I feel committed to the flag and other national symbols. 3.20 (1.77) 4.01 (1.70) 4.03 (1.81)The flag and other national symbols are inviolable. 3.94 (1.84) 4.56 (1.77) 4.38 (1.86)

Table S5. Means and Standard Deviations in Study 2. Instrumenta

l UseValue

Consistent ControlM (SD) M (SD) M (SD)

Sacred 3.93 (1.77) 4.76 (1.63) 4.64 (1.57)Should not be compromised, no matter the benefits (money or otherwise) 4.75 (1.61) 5.14 (1.54) 5.07 (1.53)I revere, respect it 4.86 (1.58) 5.33 (1.37) 5.33 (1.33)Involves principles that I would defend under any circumstances 4.65 (1.64) 5.20 (1.38) 5.10 (1.36)Cannot be contaminated by financial or other commercial interests 3.76 (1.86) 3.94 (1.80) 4.14 (1.86)

Involves issues or values I could compromise if the situation demands it (Reverse-coded) 4.15 (1.56) 4.56 (1.66) 4.00 (1.60)

38


High Prevalence Low Prevalence

Instrumental Use

M (SD)

Value-consistent

useM (SD)

Instrumental Use

M (SD)

Value-consistent use

M (SD)Sacred 3.60 (1.57) 4.20 (1.56) 3.87 (1.64) 4.12 (1.72)Should not be compromised, no matter the benefits (money or otherwise) 3.44 (1.56) 3.89 (1.60) 3.83 (1.50) 4.00 (1.75)I revere, respect it 4.13 (1.58) 4.94 (1.20) 4.64 (1.42) 5.02 (1.27)Involves principles that I would defend under any circumstances 4.28 (1.60) 4.80 (1.29) 4.64 (1.47) 4.70 (1.41)Cannot be contaminated by financial or other commercial interests 3.95 (1.56) 5.00 (1.20) 4.55 (1.44) 4.97 (1.30)

Involves issues or values I could compromise if the situation demands it (Reverse-coded) 4.31 (1.56) 5.38 (1.20) 4.87 (1.37) 5.26 (1.35)


Instrumental Use

Value Consistent

Taboo Tradeoff

M (SD) M (SD) M (SD)Sacred 3.43 (1.68) 3.94 (1.84) 4.87 (1.42)Should not be compromised, no matter the benefits (money or otherwise) 3.30 (1.49) 4.04 (1.51) 4.29 (1.59)I revere, respect it 3.63 (1.38) 4.33 (1.37) 4.53 (1.50)Involves principles that I would defend under any circumstances 4.29 (1.57) 4.47 (1.50) 4.91 (1.28)Cannot be contaminated by financial or other commercial interests 2.34 (1.24) 3.57 (1.60) 4.14 (1.86)

Involves issues or values I could compromise if the situation demands it (Reverse-coded) 4.30 (1.62) 4.08 (1.16) 4.31 (1.44)

39


Instrumental Support

Instrumental Threat

Value Consistent

Use Taboo TradeoffM (SD) M (SD) M (SD) M (SD)

Sacred 4.02 (1.78) 4.32 (1.70) 4.45 (1.73) 4.78 (1.75)Should not be compromised, no matter the benefits (money or otherwise) 4.80 (1.64) 5.22 (1.66) 5.09 (1.60) 5.58 (1.51)I revere, respect it 5.23 (1.41) 5.40 (1.41) 5.61 (1.27) 5.58 (1.41)Involves principles that I would defend under any circumstances 4.64 (1.53) 4.95 (1.50) 4.90 (1.45) 5.21 (1.46)Cannot be contaminated by financial or other commercial interests 3.79 (1.72) 4.20 (1.92) 4.14 (1.86) 4.62 (1.94)

Involves issues or values I could compromise if the situation demands it (Reverse-coded) 4.48 (1.69) 4.99 (1.40) 4.79 (1.50) 4.95 (1.68)


Instrumental Use

Value Consistent

M (SD) M (SD)

Sacred 3.92 (1.56) 3.79 (1.64)

Should not be compromised, no matter the benefits or costs (money or otherwise) 3.89 (1.43) 4.16 (1.48)

I revere, respect it 5.24 (1.23) 5.52 (1.30)

Involves principles that I would defend under any circumstances 4.75 (1.37) 5.19 (1.40)

Cannot be compromised by financial or other commercial interests 3.01 (1.35) 3.61 (1.70)

Involves issues or values I could compromise if the situation demands it (Reverse-coded)

4.16 (1.20) 4.37 (1.47)

40

Summary of studies not included in paper

Original Study 1

Participants. Participants were 155 American adults (Mage = 33.4 years; 61% female)

recruited on Amazon’s Mechanical Turk. An additional 14 adults were recruited, but failed an

attention check (“Please skip this item—do not select an answer on this line”; Oppenheimer,

Meyvis, & Davidenko, 2009), and were excluded from the analyses according to a decision made

prior to conducting the study.

Procedure. Participants were randomly assigned to one of three conditions. In the

instrumental use condition, participants then read about an organization that used the American

flag in a recent campaign because their analyses indicated that it would be a profitable approach.

In other words, the organization used patriotism to serve their self-interest. In the value-

consistent use condition, participants likewise read about an organization that used the American

flag in a recent campaign, but their primary aim was to promote national pride. Thus, in both the

instrumental use and value-consistent use conditions, a national symbol was used, but to

different ends. Participants in the control condition read neutral text about the American flag and

then proceeded to the measures.

In all conditions, participants then completed neutral filler items (e.g., “I prefer practical

jokes to verbal humor”) to mask the hypotheses. For the key dependent measure, participants

then completed a measure of sacred values adapted from Hanselmann and Tanner (2008).

Specifically, participants indicated their level of agreement or disagreement on a scale from 1

(Strongly disagree) to 7 (Strongly agree) with five statements assessing how sacred national

symbols were perceived to be (e.g., “I believe that the flag and other national symbols are

41

sacred”; “I believe that the flag and other national symbols should never be sacrificed, no matter

what the benefits”). The items were highly correlated ( = .90), and were averaged to create a

single measure of value sacredness. Participants also completed a manipulation check regarding

the intent of the organization in the instrumental use and value-consistent use conditions (i.e.,

“The organization launched the campaign to increase profits”), and provided their demographic

information. Finally, participants completed open-ended questions examining awareness of the

hypotheses (e.g., “what do you think the researchers are testing in this study?”), and were

debriefed.

Results and Discussion

Manipulation Check. Participants in the instrumental use condition were significantly

more likely to report that the campaign was intended to yield profits (M = 6.32, SD = 1.07) than

were those in the value-consistent use condition (M = 4.08, SD = 1.79), t(102) = 7.80, p < .001, d

= 1.53, indicating that the manipulation was successful.

Perceived Sacredness. Perceptions of sacredness differed significantly by condition,

F(2, 152) = 4.27, p = .016, ɳ2 = .053. Pairwise comparisons revealed that participants perceived

national symbols to be less sacred in the instrumental use condition (M = 3.82, SD = 1.63) than

in either the value-consistent use condition (M = 4.61, SD = 1.45), t(102) = -2.60, p = .011, d =

0.51, or the control condition (M = 4.52, SD = 1.35), t(102) = -2.32, p = .022, d = 0.45. No

significant difference in perceived sacredness emerged between the value-consistent use and

control conditions, t(100) = 0.30, p = .769, d = 0.06.

Original Study 2

Experimental Materials:

Please see https://osf.io/mfpzx/?view_only=4df1d286d00443bd92431cb4d7cfbe00 for tweet images.

https://osf.io/mfpzx/?view_only=4df1d286d00443bd92431cb4d7cfbe00

42

Please indicate your agreement with the following items, according to how you feel right now, at this point in time (on a scale from 1 = not at all to 7 = very much so). I believe that environmentalism is…

Pure Tainted (reverse-coded) Clean Sacred Should not be sacrificed, no matter the benefits Involves principles that I would defend under any circumstances

Method

Participants. Participants were 223 American adults (Mage = 33.4 years; 64% female)

recruited online via MTurk on Earth Day. An additional 12 adults were recruited, but failed an

attention check, and an additional 5 did not complete the study, and were excluded from the

analyses according to a decision made prior to conducting the study.

Procedure. Participants were invited to participate in a study ostensibly designed to

assess their beliefs and attitudes. Participants were randomly assigned to one of three conditions:

instrumental use vs value-consistent use vs baseline control. In the instrumental and value-

consistent use conditions, participants were first asked to evaluate Earth Day campaigns.

Participants in the instrumental use condition viewed a real campaign pretested1 to be high on

perceived instrumentality (a tweet from Nascar reading “Happy Earth Day”), while participants

in the value-consistent use condition viewed a campaign pretested to be low on perceived

1 To identify real Earth Day campaigns associated with instrumentality, 49 participants rated, on 7-point scales, a series of Earth Day campaigns (tweets) on the extent to which they were associated an intention to “primarily increase profits”. The campaign selected as instrumental was associated with significantly higher intentions to pursue profits (M = 5.98, SD = 1.09) than was the campaign selected as the value-consistent campaign (M = 2.51, SD = 1.31), paired t(48) = 9.10, p < .001. These campaigns did not significantly differ in terms of engagement, ease of reading, and aesthetic appeal, ps > .376. Note that the images were somewhat altered to increase consistency between conditions.

43

instrumentality (a tweet from Conservation International also reading “Happy Earth Day”) (see

Supplemental Materials). Control participants did not view a campaign.

After viewing the campaign, participants completed filler items about the aesthetics of

the campaign (e.g., “The image was engaging”). Participants then completed the six item

Sacredness Scale and the manipulation check. Participants also indicated their age, gender,

ethnicity, education level, Earth Day participation (“How involved have you been in Earth Day

activities this year? And “[Outside of this study] how exposed have you been to Earth Day

campaigns?), and political orientation (on a scale from 1 = Extremely liberal to 5 = Extremely

conservative). Finally, participants also answered the open-ended question, “What do you think

the researchers are testing in this study?” No participants indicated awareness of our specific

research hypotheses.

Measures

Manipulation Checks. Participants indicated their agreement or disagreement with the

statements “The campaign primarily intended to increase profits” and “The campaign primarily

intended to support the environmental cause” on scales from 1 (Strongly Disagree) to 7

(Strongly Agree).

Sacredness. To assess value sacredness, we adapted a measure from Chakroff, Dungan,

and Young (2013). Participants were asked to rate their beliefs about the value of

environmentalism along six dimensions (I believe that environmentalism is “pure,” “tainted”

(reverse-coded), “clean,” “sacred,” “should not be sacrificed, no matter the benefits”; “involves

principles that I would defend under any circumstances”; α = .85).


44

Examining the critical dependent measure, an ANOVA examining the effect of condition

on sacredness revealed a significant effect of condition, F(2, 220) = 5.94, p = .003, ɳ2 = .051.

Participants in the instrumental use condition (M = 5.18, SD = 1.16) rated environmentalism as

significantly less sacred than did those in the value-consistent (M = 5.71, SD = 0.96) and control

conditions (M = 5.65, SD = 1.04), ts > -2.55, ps < .012, ds > 0.40. The value-consistent and

control conditions did not significantly differ from each other (p = .778).

In analyzing the control variables, only political orientation significantly predicted

sacredness, b = -.22, p = .005, with more conservative participants holding environmental

protection as less sacred. This finding is consistent with past research (e.g., Dunlap, Xiao, &

McCright, 2001; Feinberg & Willer, 2013). Similarly, engagement with Earth Day activities

predicted sacredness, b = .27, p < .001, with more engagement predicting greater sacredness.

Because neither variable significantly interacted with the predictor variable, nor did controlling

for this variable affect the results, these variables are not discussed further.

Study 2 again found support for the value corruption hypothesis: Participants who

observed a real Earth Day campaign pretested to score high on instrumentality subsequently

viewed environmentalism as less sacred compared with participants who viewed an Earth Day

message rated low on instrumentality or participants who observed no message.

45

Study 8

Study 8 had three main goals. The first was to test whether mere contemplation of

instrumental use would be sufficient to demonstrate the current effect (cf., Tetlock et al., 2000).

In Study 8, we sought to compare this values activation account with our account by exploring

the effects of instrumental use on value sacredness before versus after it occurs. In both cases,

the instrumental motive is active. However, we predict corruption effects will only occur after

the sacred value has been used in an instrumental manner, when the violation of the value’s

sacred status has occurred. In the case of mere contemplation, the value has not yet been tied to

instrumental aims (e.g., Baron & Leshner, 2000; Baron & Spranca, 1997; Lichtenstein, Gregory,

Irwin, 2007; Mandel & Vartanian, 2008). As a second aim, we sought to rule out possible

alternative explanations based on demand characteristics. To do so, participants completed an

implicit measure of sacredness in addition to the explicit measures. Finally, we assessed an

important behavioral outcome: participants’ willingness to donate to value-relevant (or less

relevant) charities (Stikvoort, Lindahl, & Daw, 2016; Tetlock et al., 2000). If the corruption

effect is specific to the value that has been used instrumentally, participants’ general willingness

to support values-driven causes will not be affected; rather, they will be less willing to donate to

causes relevant to the corrupted value.

To test these predictions, participants were randomly assigned to 1 of 3 conditions in a

between-subjects design: instrumental use vs. contemplated instrumental use vs. value-consistent

use. After reading about an environmental campaign promoted for instrumental or value-

consistent reasons, participants completed the measures of value sacredness and their willingness

to donate to an environmental cause or other causes.

Method

46

Participants. Participants were 128 undergraduates at a Midwestern university (Mage =

20.59 years; 59% female). We sought to collect 150 participants (i.e., 50 per cell), but some

participants did not show up to their appointments, and we ended data collection at the end of the

final academic quarter for the year. A power analysis (using G*Power 3.1) indicated that this

final sample size gave us 71% power to detect a medium effect. Given the reduced power of this

study, it is presented in full only here in the Supplemental Online Materials.

Procedure. Upon arrival at the laboratory, participants first read initial instructions

indicating that they were participating in a study assessing evaluations and recall of various

communications. In all conditions, they then read a transcript between a CEO and a board of

directors. In the instrumental-use condition participants read the following scenario:

The CEO of a company spoke to the company’s board of directors and said, “We are

thinking of starting a new environmental campaign. The chairman of the board replied,

“Good, green sells. Let’s make as much profit off of this as we can.” After deliberating

the issue, the board members all agreed, and the company went forward with the

environmental campaign.

The contemplated instrumental-use condition was similar, except that instead of going

forward with the campaign, participants read that “the board members will deliberate the issue.”

In the value-consistent use condition, participants read a similar dialogue, but chairman of the

board instead replied, "Good, let's increase our commitment the environmental cause as much as

we can."

Participants then completed the implicit measure of the purity component of sacredness.

This measure draws on the tendency for purity (impurity) to be grounded in the perceptual

experience of the color white (black) (Sherman & Clore, 2009; Sherman, Haidt, & Clore, 2012).

47

Specifically, under the guise of a study on image recall task, participants were exposed to a series

of neutral (e.g., “table,” “forecast”) and value-relevant words (e.g., “green,” “environmental”) in

fonts varying in color on the white-black spectrum (adapted from Meier, Robinson, Crawford, &

Ahlvers, 2007). After viewing each word, participants viewed a blank screen for 11 seconds

before being asked to indicate the shade of the previous word on a 5-point gray shades scale. The

degree to which participants associated environmentalism with darker as opposed to lighter

shades served as an index of perceived impurity.

After completing the implicit measure, participants completed explicit measures of the

perceived sacredness of environmentalism (as in the previous studies), and of other values

(fairness, health, social order, pleasure, and equality). Participants then completed the

manipulation checks and demographic measures. To assess donation behavior, participants were

told that they have been provided with an additional $3, which they could either keep or donate

part or all of to a charity. Participants then indicated their desired donation amount (from $0 to

$3), and could choose from one of three charities. Each charity captured a different sacred value:

Wildlife Conservation International (environmental protection), Get Fit Foundation (health/well-

being), and Teaching Tolerance (diversity).


Implicit measure. Consistent with Meier et al., we analyzed participants’ number of

correct choices by word color. To the degree that participants associate a given concept with

darkness rather than lightness, they will be less accurate on trials in which the concept appeared

in a lighter font. Participants’ lightness-matching judgments were accurate on average 43.16%

(SD = 15.77%) of trials (chance accuracy = 20%). Thus, although the task was challenging,

perceptual judgments were more accurate than chance, t(127) = 16.62, p < .001, d = 2.95.

48

Validating use of this measure, performance on the light and medium trials was significantly

correlated with the explicit measure of sacredness, rs > .27, ps < .002. Performance on the dark

trials was not, r = .00, p = .99.

There was a main effect of condition on the correct categorization of light trials, F(2,

125) = 3.41, p = .036, ɳ2 = .052. Participants in the instrumental-use condition answered

significantly fewer questions correctly (M = 0.81, SD = 0.82) than did those in the contemplation

condition (M = 1.40, SD = 1.19), t(83) = -2.67, p = .009, d = 0.58, and marginally fewer than did

those in the in the value-consistent use condition (M = 1.18, SD = 1.11), t(84) = -1.76, p = .083,

d = 0.38. The contemplation and value-consistent conditions did not significantly differ (p

= .385). There was also a main effect of condition on the correct categorization of medium trials,

F(2, 125) = 3.71, p = .027, ɳ2 = .056. Participants in the instrumental-use condition answered

significantly fewer questions correctly (M = 1.00, SD = 0.87) than did those in the contemplation

condition (M = 1.50, SD = 0.99), t(83) = -2.47, p = .016, d = 0.53, and significantly fewer than

did those in the in the value-consistent use condition (M = 1.52, SD = 1.10), t(84) = -2.39, p

= .019, d = 0.51. The contemplation and value-consistent conditions did not significantly differ

(p = .959). No significant differences emerged on the dark trials, or for the neutral words.

Explicit ratings of sacredness. There was a main effect of condition on ratings of value

sacredness, F(2, 125) = 17.02, p < .001, ɳ2 = .214. Participants in the instrumental-use condition

reporting holding environmentalism as significantly less sacred (M = 3.07, SD = 0.98) than did

those in the value-consistent use condition (M = 4.22, SD = 0.96), t(84) = -5.46, p < .001, d =

1.18, or in the instrumental-use contemplation condition (M = 4.00, SD = 0.96), t(83) = -4.44, p <

.001, d = 0.96. Participants in the value-consistent use and control conditions did not differ

49

significantly (p = .562, d = 0.22). Participants’ ratings of the sacredness of other values (fairness,

health, social order, pleasure, and equality) did not differ by condition (ps > .332).

Donation behavior. Because the data were right skewed and contained many zeroes, we

analyzed whether condition predicted whether or not participants donated to the environmental

cause. We conducted a binary logistic regression in which we regressed donation (0 = did not

donate to the environmental cause; 1 = did donate) onto two dummy coded variables (1 =

instrumental use; 0 = mere contemplation; 0 = value-consistent and 0 = instrumental use; 1 =

mere contemplation; 0 = value-consistent). The results revealed the predicted effect of

instrumental use condition, b = -2.41, χ2 (2, N = 128) = 4.98, p = .026, such that participants in

the instrumental use condition were significantly less likely to donate to the environmental cause

than were those in the other two conditions. While only 2% of participants in the instrumental-

use condition donated to the environmental cause, 24% did so in the contemplated instrumental

use condition, and 26% did so in the value-consistent use condition. There was, however, no

effect of the contemplated condition on the tendency to donate to the environmental cause, b = -

0.12, χ2(2, N = 128) = 0.05, p = .828.

Importantly, there was no effect of instrumental use condition, b = -0.43, χ2(2, N = 128) =

0.95, p = .331, or contemplated condition, b = 0.00, χ2(2, N = 128) = 0.00, p = 1.00, in predicting

the tendency to donate in general. Thirty-four percent of participants in the instrumental use

condition donated, and 45% of participants in both the contemplated and value-consistent

conditions donated. This helps to rule out the possibility that instrumentality simply inhibits

prosocial behavior in general.

Mediation. To explore whether perceptions of sacredness drove the effect of the value

use condition on donating to the environmental cause, we used Preacher and Hayes' (2008)

50

PROCESS Model 4 for mediation. The results revealed a significant indirect effect of sacredness

in the mediated moderation model, b = -2.22, 95% CI [-4.29, -.15], indicating significant

mediation.

Study 8 again found support for the value corruption hypothesis: Participants who

observed the instrumental use of environmentalism subsequently viewed that value as less

sacred, as captured by both implicit and explicit measures, and were less willing to donate to the

environmentalism cause. Moreover, the instrumental use had to have taken place for the value

corruption effect to occur, suggesting that merely activating a motivationally incompatible value

does not reduce the perceived sacredness of sacred values.

Study 9

In Study 9, we examined whether participants who witnessed a value being used for

instrumental gain would subsequently be more willing to endorse ambiguous uses of the value as

being value-consistent. One key marker of the sacred is people’s unwillingness to blend elements

of the secular with the sacred, and exert effort to restrict the meaning of that sacred object or

value (Belk et al., 1989; Ginges & Atran, 2009; Tetlock, 2002, 2003). To the extent that a value

becomes less sacred, people may hold less stringent criteria for what comprises a value-

consistent use. To test this idea, participants completed the instrumentality manipulation, and

then rated three organizations which were pretested to vary in people’s willingness to describe

them as pro-environmental. Participants indicated the degree to which they believed the

organizations were pro-environmental.

Method

Pretest. To select the stimuli for Study 9, 58 participants recruited on Amazon’s

Mechanical Turk rated the degree of pro-environmentalism (on a scale from 1 = not at all to 7 =

51

a great deal) of ten organizations that have made “green” claims, but range in the degree to

which they might be perceived as driven by environmentalism (versus profit). On the basis of

these results, we selected three organizations that ranged in ratings of pro-environmentalism for

use in Study 8. The less pro-environmental organization (Toyota, an automotive manufacturer)

was rated as significantly less pro-environmental (M = 4.10, SD = 1.60) than was the moderately

pro-environmental organization (West Paw, a sustainable pet store) (M = 5.55, SD = 1.26), t(57)

= -6.46, p < .001, d = -1.71, or the highly pro-environmental organization (Conservation

International, a nonprofit environmental organization), t(57) = -8.67, p < .001, d = -2.30. The

moderately environmental organization was rated as less pro-environmental than was the highly

pro-environmental organization, t(57) = -4.66, p < .001, d = -1.22.

Participants. Participants were 142 American adults recruited on Amazon’s Mechanical

Turk. An additional 23 participants were recruited, but failed an attention check asking them to

correctly select which type of task they would encounter in the study, and were excluded from

the analyses according to a decision made prior to conducting the study. A power analysis (using

G*Power 3.1) indicated that this left us with 75% power to detect a medium-sized effect. As with

Study 8, we report this study in full only in the Supplemental Materials because of this reduced

power.

Procedure. Participants were randomly assigned to 1 of 3 conditions (instrumental use

vs. value-consistent use vs. pure instrumentality) in a between-subjects design. Participants were

informed that they would participate in a study about interpersonal communication, and would

view a message that was written by another individual named Jamie. In all conditions, Jamie was

soliciting donation requests. In the instrumental use condition, he was a being paid $0.25 for

every $1 collected. In the value-consistent condition, Jamie was a volunteer, collecting donations

52

for the Wildlife Conservation Network, supporting the environmentalism cause. In the pure

instrumentality condition, Jamie was working for Accel Innovations for profit. The pure

instrumentality condition was again included to rule out the possibility that mere exposure to

self-interested behavior might make participants less concerned about the sacredness of the

environmental cause (e.g., Evans, Maio, Corner, Hodgetts, Ahmed, & Hahn, 2013).

After completing filler items, participants were asked to rate each of the three

organizations from the pretest in a randomly generated order. For each, participants completed a

5-item scale designed to assess the degree to which the organizations were rated as pro-

environmental (e.g., “To what extent would you rate this organization as pro-environmental?”)

on a scale from 1 (Not at all) to 7 (A great deal). Participants then completed a manipulation

check regarding the intent of the organization in the instrumental use and value-consistent use

conditions (i.e., “Jamie is seeking to increase his profits”), and completed demographic

measures. Participants also answered the open-ended question, “What do you think the

researchers are testing in this study?”

Results

Manipulation Check. There was a main effect of condition on perceived profit

motivation, F(2, 139) = 113.01, p < .001, 2 = .619. Participants in the instrumental use

condition (M = 6.04, SD = 1.38) and in the pure instrumentality conditions (M = 6.38, SD = 1.00)

were significantly more likely to report that Jamie was motivated by profit than were those in the

value-consistent use condition (M = 2.43, SD = 1.77), t(93) = 12.43, p < .001, d = 2.58, and t(93)

= 13.44, p < .001, d = 2.79, respectively. The instrumental use and pure instrumentality

conditions did not significantly differ, t(93) = 1.34, p = .489, d = 0.28. Participants in the

instrumental use conditions were significantly more likely to report that the campaign was

53

intended to yield profits (M = 6.14, SD = 0.78) than were those in the value-consistent use

conditions (M = 2.45, SD = 1.56), t(115) = 16.25, p < .001, d = 2.95.

Pro-environmental Ratings. There was a main effect of condition on ratings of the less

pro-environmental organization, F(2, 139) = 7.01, p = .001, 2 = .092, but no main effect of

condition on the ratings of the moderately F(2, 139) = 0.64, p = .527, 2 = .009, or highly pro-

environmental organizations, F(2, 139) = 0.989, p = .379, 2 = .014. For the less pro-

environmental organization, participants in the instrumental use condition were significantly

more likely to rate the organization as pro-environmental than were participants in the value-

consistent use condition, t(93) = 3.39, p = .003, d = 0.70, and the pure instrumentality condition,

t(93) = 2.57 = .030, d = 0.53. There were no significant differences between the value-consistent

and pure instrumentality conditions, t(93) = 0.83, p = .684, d = 0.17.

Study 9 again finds support for the value corruption hypothesis, this time using a

different indicator of value sacredness: a relaxing of standards for what constitutes value-

consistent use. However, given the reduced power of this study, we interpret these effects with

caution.

54

Power Analyses for the Experiments

Study 1

55

Study 2

56

Study 3

57

Study 4

58

Study 5

59

Study 6

Documents

supp.apa.org · Web view2021. 4. 30. · Let’s go forward with this new line. ... .039. 6. 2 cells. Difference of means. t (277) = 2.68, p = .008. ... Let’s make as much profit