Delineating a method to study cross-cultural differences with experimental control: The voice effect and countercultural contexts regarding power distance

Journal of Experimental Social Psychology 49 (2013) 624–634

Contents lists available at SciVerse ScienceDirect

Journal of Experimental Social Psychology

j ourna l homepage: www.e lsev ie r .com/ locate / jesp

Delineating a method to study cross-cultural differences with experimental control:The voice effect and countercultural contexts regarding power distance

Kees van den Bos a,⁎,1, Joel Brockner b,1, Michael van den Oudenalder a,Shanmukh V. Kamble c, Afreen Nasabi c

a Utrecht University, The Netherlandsb Columbia University, USAc Karnatak University, India

H I G H L I G H T S

► We propose a method to study cross-cultural differences with experimental control.► This method uses countercultural conditions to study cross-cultural differences.► We examine reactions from high and low power distance participants to voice opportunities.► Experimental conditions remind about countercultural values regarding power distance.► Cultural and situational high (vs. low) power distance yields weaker reactions to voice.

⁎ Corresponding author at: Department of Social PsychBox 80140, 3508 TC Utrecht, The Netherlands.

E-mail address: [email protected] (K. van den Bos1 Kees van den Bos and Joel Brockner contributed eq

Allan Lind for his comments on an earlier version of thi

0022-1031/$ – see front matter © 2013 Elsevier Inc. Allhttp://dx.doi.org/10.1016/j.jesp.2013.02.005

a b s t r a c t
a r t i c l e i n f o
Article history:Received 4 February 2012Revised 9 January 2013Available online 19 February 2013

Keywords:Cross-cultural differencesExperimental methodCountercultural contextsVoicePower distance

This paper proposes a method to study cross-cultural differences with experimental control. We illustratethis method by examining how participants from India (a high power distance culture) and the Netherlands(a low power distance culture) react to being allowed or denied an opportunity to voice their opinions. Weargue that one way to evaluate the influence of the assumed cultural differences in power distance is to as-sign participants to conditions that elicit “countercultural” psychological states, that is, conditions that primelow power distance in India and high power distance in the Netherlands. To the extent that the results in thecountercultural (experimental) conditions meaningfully differ from those observed in the control conditionsin which no values are emphasized explicitly, we gain insight into the psychological dimensions that accountfor cross-cultural differences in people's reactions. The findings presented indeed suggest that the random as-signment of participants to countercultural conditions provides cross-cultural researchers with a powerfultool to examine the causal impact of meaningful psychological dimensions that are presumed to differ acrosscultures. The results further reveal that when high power distance was emphasized either because of nationalculture or situational cues, participants showed less strong reactions to the voice versus no-voice manipula-tion than when low power distance was emphasized as a result of either national culture or situational cues.Implications and limitations of this countercultural-experimental approach for the study of cross-cultural dif-ferences as well as the psychology of voice and power distance are discussed.

© 2013 Elsevier Inc. All rights reserved.

Introduction

People are cultural animals (Baumeister, 2005) and are influencedheavily by their cultural surroundings (e.g., Cohen & Leung, 2009;Fiske, 2006; Hofstede, 2001; Leung, 2005; Markus & Kitayama, 1991;Martin, 1999; Schaller & Crandall, 2004). It does not come as a sur-prise, therefore, that many social psychologists are interested in,

ology, Utrecht University, P. O.

).ually to this article. We thanks paper.

rights reserved.

indeed fascinated by cross-cultural differences. As a result, many im-portant advances have been made in this field of inquiry (for recentreviews, see, e.g., Chiu & Hong, 2007; Heine, 2010). For example,cross-national research has enabled researchers to evaluate thecross-cultural generality of their findings. Many social phenomenahave been demonstrated in Western countries which raises the im-portant question whether these theories apply in non-Western con-texts (Heine, 2010; Hofstede, 2001). Indeed, research has shownthat some of the most important and robust phenomena inWestern-based social psychology do not always emerge or do notemerge to the same degree in non-Western countries (e.g., Brockner,2003; Leung, 2005; Markus & Kitayama, 1991).

http://dx.doi.org/10.1016/j.jesp.2013.02.005

mailto:[email protected]

http://dx.doi.org/10.1016/j.jesp.2013.02.005

http://www.sciencedirect.com/science/journal/00221031

625K. van den Bos et al. / Journal of Experimental Social Psychology 49 (2013) 624–634

Complementing cross-cultural researchwith experimental control

We argue here that an important concern when studying cross-cultural differences is the need to have experimental control. Thatis, often cross-cultural research examines people from two or morecountries and assumes that the participants differ along certain psy-chological dimensions, which, in turn, elicit differences in dependentvariables of cognition, affect, or behavior. If the research findings sug-gest that the participants in the different countries react differently,and that these reactions are in accordance with the psychological di-mensions presumed to differ across cultures, it is still necessary toprovide evidence that the culturally-varying reactions can be attrib-uted to the hypothesized psychological dimensions. For example, ifpeople from cultures that score high on a psychological dimensionsuch as power distance (see, e.g., Hofstede, 2001) react differently to-ward issues such as opportunities to voice their opinions than peoplefrom cultures that score low on the dimension (Brockner et al., 2001),then researchers still need to provide evidence as to why these effectsemerged.

Sometimes researchers try to provide such evidence by measuringthe relevant psychological dimensions (Brockner, 2003), and thenconducting tests of mediation to see whether the cultural differenceswere attributable to the psychological dimensions. However, measur-ing the hypothesizedmediating psychological dimensions may be dif-ficult (see, e.g., Spencer, Zanna, & Fong, 2005). For example, thepsychological dimensions may reflect unconscious processes thatare difficult to assess. Relatedly, measuring psychological dimensionsin cross-cultural research typically involves self-reports, which maybe problematic because of people's unwillingness or inability to dis-close their true beliefs (e.g., Nisbett & Ross, 1980). Furthermore, awell-known issue in cross-cultural studies is that measures of thesame construct may not mean the same thing to people from differentcultures (Hui & Triandis, 1985). In short, there may be importantproblems associated with measuring the psychological dimensionspresumed to account for cross-cultural differences.

In addition, evenwhen it is possible tomeasure reliably the hypoth-esized psychological mediators of the effects of culture, the ensuing sta-tistical analyses can at best provide only correlational evidence whichlimits empirical strength of the study in question. This concern is cer-tainly not confined to cross-cultural research. Rather, it is inherent topractically any study in which the independent or intervening variablesare measured rather than manipulated (Spencer et al., 2005).

In the present paper we argue that there is a need incross-cultural research to not only provide evidence of why peoplefrom different cultures show different reactions, but also to do so inways that entail a high degree of internal validity. Therefore, we sug-gest that it is important to complement extant cross-cultural inqui-ries with studies that provide experimental control over thepsychological dimensions hypothesized to account for differencesbetween cultures. We further argue that for purposes of internalvalidity it is better to use an experimental approach in which partic-ipants are randomly assigned to conditions differing along thedimension hypothesized to account for the results. Indeed, one ofthe main virtues of such an approach is that it is likely to hold con-stant all factors other than the construct being manipulated. As a re-sult, differences that emerge on the dependent variable are likely tohave been causally influenced by the manipulated construct. In thecurrent paper we illustrate such an experimental approach tocross-cultural research on how people from cultures that are eitherhigh or low in power distance react to being allowed or denied an op-portunity to voice their opinions.

An experimental approach to cross-cultural differences

Of course, in cross-cultural research it is not possible to randomlyassign participants to different cultures. However, it is possible to

randomly assign people from different cultures to either a controlcondition in which nothing is done and in which people hence arelikely to default to the values and beliefs that are predominant totheir culture or to an experimental condition which primes valuesand beliefs that are directly contrary to the default ones in a given cul-ture. If differences emerge on the dependent variable when thismethodology is used it is quite likely that they were caused by theconstruct inherent to the experimental manipulation.

We further note that a basic notion in experimental social psy-chology dictates that it is easier to interpret research findings whenthe results found in control conditions can be contrasted with thefindings obtained in the experimental conditions. In other words, asignificant difference between control and experimental conditionsis easier to interpret than a null effect. We therefore propose toprime participants in our experimental conditions with values thatare in direct contrast with the default ones in their culture. If thevalues in the experimental conditions are indeed counter to what isdefault in the culture that is studied, then priming participants withcountercultural beliefs should lead to reliably different effects on par-ticipants' reactions to subsequent events, compared to participants'reactions in control conditions in which no explicit values havebeen primed. Moreover, the internal validity of such findings wouldbe much higher than if the independent variable had been measuredrather than manipulated.

In short, we propose an experimental research method to studycross-cultural differences that involves the following features:(1) studying participants from cultures that differ in ways thattheory suggests will produce meaningful effects on importanthuman reactions (see, e.g., Chiu & Hong, 2007; Cohen & Leung,2009; Heine, 2010; Hofstede, 2001; Hui & Triandis, 1985; Leung,2005; Markus & Kitayama, 1991), and (2) incorporating the conceptof a countercultural psychological state in a controlled experimentaldesign.

Initial support for the value of the experimental method we are ad-vocating was provided in a recent study by Van den Bos et al. (2010).These authors took as their point of departure the results of an earlierstudy by Brockner et al. (1998) that examined the interactive effect ofgiving people voice and their perceived competence to provide mean-ingful input on their satisfaction with the decision-making process.The Brockner et al. (1998) study showed that the tendency for peopleto be more satisfied when they were given voice was significantamong those who believed that they could provide meaningful input.In contrast, the satisfaction level of those who saw themselves as lesscapable of providing meaningful input was relatively unaffected bytheir level of voice.

Brockner et al. (1998) conducted their research in the UnitedStates in which people have been shown to emphasize competitionand achievement, or what Hofstede (2001) has called “masculinity.”Van den Bos et al. (2010) hypothesized that in cultures valuing nur-turance and equality (“femininity”) the results would look rather dif-ferent: in this instance those who saw themselves as less capable ofproviding meaningful input were expected to be more motivated tohave voice and, as a result, their satisfaction level would be more pos-itively affected by having voice relative to their counterparts who sawthemselves as more capable of providing input.

To test their hypothesis Van den Bos et al. (2010) studied partici-pants from both the Netherlands (a culture high in femininity) andthe United States (a culture high in masculinity). At random, half ofthem were induced into a countercultural psychological state (femi-ninity for the Americans andmasculinity for the Dutch)whereas noth-ing was done to the other half, in which participants' default culturalvalues (masculinity for the Americans and femininity for the Dutch)were expected to be more influential. As predicted, Americans whowere experimentally induced to be feminine showed similar reactionsas the Dutch to whom nothing was done: in both instances voicewas more positively related to satisfaction among those who saw

626 K. van den Bos et al. / Journal of Experimental Social Psychology 49 (2013) 624–634

themselves as less capable of providing input. Furthermore, the Dutchwhowere experimentally induced to bemasculine showed similar re-actions to the Americans to whom nothing was done: in both in-stances voice was more positively related to satisfaction amongthose who saw themselves as more capable of providing input.

Whereas the Van den Bos et al. (2010) findings provided prelimi-nary evidence that their experimental method offered a promisingway to examine cross-cultural differenceswith a high degree of internalvalidity, further research is clearly needed. For instance, it is importantto evaluate the method used by Van den Bos et al.: (1) using othercultural dimensions besides masculinity–femininity, (2) when examin-ing other important conceptual and empirical questions besides howsatisfaction is interactively influenced by voice and people's perceivedcapabilities to provide meaningful input, and (3) when doing a culturalcomparison other than one between the United States versus theNetherlands. Accordingly, in the present study we used the experimen-tal method of Van den Bos et al. to evaluate whether differences in thecultural dimension of power distance (Hofstede, 2001) moderated theeffects of the presence versus the absence of voice on how participantsfrom India and the Netherlands felt they were treated during adecision-making process.

Voice and power distance

In testing the experimental approach to cross-cultural differ-ences we focus on how people in cultures or contexts that value ei-ther high or low distance to individuals or authorities who havepower over them respond to the presence versus the absence of op-portunities to voice their opinions about decisions to be made. Inparticular, we examine the well-established tendency for peopleto be more satisfied with a decision when they have received, asopposed to have been denied, voice. This instance of the “voiceeffect” is probably one of the most generally accepted and best-documented findings in the procedural justice literature (e.g.,Brockner et al., 1998; Folger, 1977; Lind & Tyler, 1988; Van denBos, 2005). Given the centrality of voice in procedural justice andin many real-world situations (see, e.g., Lind, Kanfer, & Earley,1990; Thibaut & Walker, 1975; Tyler, 1987; Tyler, Rasinski, &Spodick, 1985), it is both theoretically and practically important tounderstand the conditions under which the presence or absenceof voice is more versus less likely to influence individuals' reactionsto decisions (Van den Bos, 1999). In the present study, we focus onimportant and heretofore unexamined countercultural contextsthat may have a moderating influence on the voice effect, and inso doing contribute to a better understanding of the psychology ofvoice and procedural justice.

Within the broader literature on intercultural fairness effects (see,e.g., Ang, Van Dyne, & Begley, 2003; Fischer & Smith, 2003; Francesco& Chen, 2004; Leung, Su, & Morris, 2001; Yang, Mossholder, & Peng,2009), previous theory and research have considered possible cultur-al differences in how people react to voice. Some have argued thatvoice and procedural justice effects are largely invariant across differ-ent cultures (e.g., Lind, Tyler, & Huo, 1997). Others have argued thatthe voice effect is moderated by differences between individualisticand collectivistic cultures (e.g., Leung, 2005), masculine and femininecultures (e.g., Van den Bos et al., 2010), or by cultures varying inpower distance (e.g., Brockner et al., 2001; Yang, Mossholder, &Peng, 2007; Zhang & Begley, 2011).

Power distance refers to cultural conceptions regarding the degreeof power that authorities should have over subordinates (e.g.,Hofstede, 2001; Kirkman, Chen, Farh, Chen, & Lowe, 2009; Kirkman& Shapiro, 2001). Cultures in which it is believed that superiorsshould have a great degree of power over subordinates are consid-ered to be high in power distance and cultures in which it is believedthat a small degree of power is appropriate are considered to be lowin power distance. Put differently, power distance refers to the extent

to which inequality among persons in different positions of formalpower is viewed as a natural and even desirable aspect of the socialorder (Brockner et al., 2001; Hofstede, 2001; Kirkman et al., 2009;Kirkman & Shapiro, 2001). The norms of high power distance cultureslegitimize differences in decision-making power between those whoare in high power positions versus those who are in low power posi-tions. In contrast, the norms of low power distance cultures reducepower differences among people in positions of varying levels of for-mal decision-making power.

When studying how people respond to the presence versus ab-sence of voice, power distance is an important variable because it re-fers to the extent to which people accept and expect power to bedistributed unequally. In other words, the concept of power distancerefers to the acceptance and expectation of inequality in society andone's culture. Because inequality is at the heart of many fairness is-sues (e.g., Messick, 1993) and because the psychology of voice consti-tutes an important component of procedural fairness (Brockner et al.,2001; Folger, Rosenfield, Grove, & Corkran, 1979; Tyler et al., 1985;Van den Bos, 2005), studying the possible influence of differences inpower distance on how people respond to voice and no-voice proce-dures, may contribute to our understanding of the psychology of thevoice effect.

In particular, we note that in cultures or contexts in which there isa small distance to power holders, people may consider it more ap-propriate when they receive voice and less appropriate when theyare denied voice, compared to cultures or contexts in which there isa large distance to power holders. As a result, people from highpower distance cultures expect or want voice to a lesser extent thanthose from low power distance cultures (Brockner et al., 2001).Furthermore, Brockner et al. (2001, Study 4) found in a single culturalsetting among Chinese employees that individual differences inpower distance beliefs moderated the voice effect such that thoserespondents who believed more strongly in high power distanceshowed less strong voice effects on job satisfaction, perceived organi-zational commitment, and intention to remain working for theorganization.

Other studies have found similar weakened voice or proceduraljustice effects in high as opposed to low power distance cultures(see, e.g., Lam, Schaubroeck, & Aryee, 2002; Lee, Pillutla, & Law,2000; Tyler, Lind, & Huo, 2000; Yang et al., 2007; Zhang & Begley,2011). Furthermore, prior theory and research in single-culture stud-ies have shown that people have stronger reactions to being grantedversus denied voice when, beforehand, they expected or wanted tohave voice (Van den Bos, Vermunt, & Wilke, 1996). All of this leadus to predict that cultural and contextual differences in power dis-tance should have a moderating influence on how people respondto voice versus no-voice procedures. Specifically, the literaturereviewed briefly here led us to expect that the voice effect shouldbe less pronounced in high power distance cultures or contexts thanin low power distance cultures or contexts.

The current research

Whereas prior research has provided evidence consistent with thenotion that people from higher power distance cultures will be lessinfluenced by the presence or absence of voice (Brockner et al.,2001), an important shortcoming of the previous studies is thatthey merely measured people's power distance beliefs. The fact thatthe hypothesized intervening variable of the between-culture differ-ence was measured rather than manipulated weakens the internalvalidity of the findings. Accordingly, in the present study we usedthe experimental approach advocated here to evaluate whether dif-ferences in the cultural dimension of power distance (Hofstede,2001; Kirkman et al., 2009; Kirkman & Shapiro, 2001) moderate theeffects of the presence or absence of voice on how people feel theyare treated during a decision-making process.


Participants came from two cultures: India (which is known tovalue high power distance) and the Netherlands (which is knownto value low power distance).2 One way to evaluate the influenceof cultural differences in power distance on how participantsrespond to voice and no-voice procedures is to randomly assignsome participants to conditions designed to elicit “countercultural”psychological states, that is, conditions that prime low power dis-tance beliefs in India and that prime high power distance beliefs inthe Netherlands. To the extent that the results in the countercultural(experimental) conditions meaningfully differ from those observedin the control conditions in which no values are emphasized explic-itly, we can gain greater insight into the psychological dimensionsthat account for cross-cultural differences in people's reactions.Thus, by including such countercultural conditions (i.e., the experi-mental conditions), we are on firmer ground in suggesting that thedifferent results expected to arise in the control conditions in theIndian and Dutch samples are attributable to differences in powerdistance beliefs.

After being randomly assigned either to experience a counter-cultural psychological state or not, participants in both cultureswere then asked to imagine they worked in an organization inwhich their manager either did or did not allow them opportuni-ties to voice their opinions. The primary dependent variable washow satisfied participants were with how they had been treatedby their managers. Following the line of reasoning outlined earli-er, in the sample from India the effect of the procedure manipula-tion (voice vs. no voice) on experienced satisfaction was expectedto be greater in the countercultural (or experimental) condition(in which low power distance beliefs were made salient) than inthe control condition (in which participants presumably woulddefault to the high power distance beliefs prevalent in theirculture). In the sample from the Netherlands, the effect of theprocedure manipulation on perceived satisfaction was expectedto be lower in the countercultural (or experimental) condition(in which high power distance beliefs were made salient) thanin the control condition (in which participants were expected todefault to the lower power distance beliefs inherent to theirculture).

In summary, the present study consisted of a 2×2×2 factorial de-sign, in which the independent variables were culture (India versusthe Netherlands), countercultural priming (present in the experi-mental condition versus absent in the control condition), and proce-dure (voice versus no voice). The predicted effects on treatmentsatisfaction were expected to manifest themselves in the form of athree-way interaction effect: Within India, the voice effect shouldbe stronger in the experimental condition (in which lower powerdistance beliefs were experimentally induced) than in the controlcondition, whereas within the Netherlands, the voice effect should

2 Specifically, in our researchwe examined Indian participants fromKarnatakUniversityin India andDutch students fromUtrechtUniversity in theNetherlands.Wepilot tested theassumption that Indian participants fromKarnatak University would score higher on pow-er distance than Dutch participants from Utrecht University. To this end, 60 students whocame from the same population as the participants but who did not take part in the mainstudy (34 from Karnatak University, 26 from Utrecht University; 30 women;Mage=23.13,SEage=0.48) filled out a 7-itemmeasure of power distance that was based on thework byKirkman and colleagues (Kirkman et al., 2009; Kirkman & Shapiro, 2001). This measureasked participants to indicate howmuch they agreed with items such as “people at higherlevels in organizations have a responsibility to make important decisions for people belowthem” and “people at lower levels in the organization should not have much power in or-ganizations.” Responses could range from not at all (1) to very much (7) and participants'answers to the seven items were averaged to form a reliable scale of power distance(alpha=.75). Results indicated that Indian participants indeed scored significantly higheron power distance (M=4.22, SE=0.18) thanDutchparticipants (M=3.40, SE=0.23), F(1,58)=8.33, pb .01, ηp

2=.13. This finding was not qualified by gender or age of participantsand lent support for our assumption of Indian participants from Karnatak University scor-ing higher on power distance than Dutch participants from Utrecht University.

be weaker in the experimental condition (in which higher powerdistance beliefs were experimentally induced) than in the controlcondition. To the extent that the predicted results emerged in thecontext of the present experimental design it can be concludedthat power distance beliefs causally influenced the expected differ-ences in the voice effect between participants from India and theNetherlands. Exploratively we assessed whether our Indian andDutch participants would react more strongly to either the voice orthe no-voice conditions as a function of the countercultural primingmanipulation.

Method

Participants and design

Two hundred and fifty students (157 women; Mage=21.57,SEage=0.19) participated in the study and were given course creditfor their participation. One hundred and twenty-one Indian studentsfrom Karnatak University in India and 129 Dutch students at UtrechtUniversity in the Netherlands participated and were randomlyassigned to one of the conditions of a 2 (countercultural priming:absent vs. present)×2 (procedure: voice vs. no voice) factorialdesign.3

Experimental procedure

Our experimental procedure was similar to that of Brockner et al.(2001, Studies 1 and 2). To rule out potential language effects, all ma-terials were written in English. Both our Indian and Dutch partici-pants were quite fluent in English and took approximately 15 minto complete the stimulus materials. Participants were informed atthe outset that the stimulus materials would be presented to themin unrelated parts. In these parts we induced the counterculturalpriming manipulation, the procedure manipulation, and measuredour dependent variables.

In the first part of the study, participants completed a two-questionexercise designed to prime or not prime them with counterculturalvalues regarding power distance. That is, in the experimental condi-tion participants read and answered two questions pertaining tocountercultural psychological states, that is, conditions that madesalient low power distance in India and high power distance in theNetherlands. In the control condition in both countries powerdistance beliefs were not primed. The questions we used in ourcountercultural priming manipulation were modeled directly afterseveral salience manipulations successfully used to prime otherconstructs in social psychology, including uncertainty salience(Van den Bos, 2001), status salience (Van Prooijen, Van den Bos, &Wilke, 2002), affective states (Van den Bos, 2003), mortalitysalience (Greenberg, Solomon, & Pyszczynski, 1997; Van den Bos &Miedema, 2000), and behavioral disinhibition (Van den Bos et al.,2011). It also was similar in nature to the procedure used by Vanden Bos et al. (2010) to prime counterculturally the dimension ofmasculinity–femininity.

The questions and instructions in the experimental condition inwhich low power distance was primed among our Indian participants

3 The Indian sample was somewhat older (M=23.22, SE=0.15) than the Dutchsample (M=20.08, SE=0.28) and there were more women in the Dutch sample (99women) than in the Indian sample (58 women), but controlling for both age and gen-der in our main analyses did not substantially affect our findings. For example, this stillyielded the predicted three-way interaction between country, countercultural priming,and procedure on our main dependent variable, participants' ratings of treatment sat-isfaction, F(1, 236)=9.14, pb .01, ηp

2=.04. Age and gender were therefore droppedfrom the analyses reported.


were derived from the definition of power distance by Hofstede(2001). Specifically, instructions were as follows:

This part of the study will focus on other people's potential to de-termine or direct your behavior. More specifically, we ask you toread some materials and answer some questions that ask you toimagine that you and those who have power over you (e.g., em-ployers, teachers, your parents, etc.) regard each other as equals.Thus, formal positions do not matter so much.

Question 1: Please describe a situation out of your own life inwhich there was a small distance between you and the personwho formally had power over you. Thus, we ask you to imagineand describe to us a situation in which you and a person who for-mally had power over you regarded each other more or less asequals. Could you briefly describe this situation to us?Question 2: Imagine there is a small distance between you and aperson who has power over you, thus that you and a personwho formally has power over you regard each other more or lessas equals. Could you briefly describe how you would feel in sucha situation and why it may be a good thing when a person withpower treats you as being equal?

Also building on Hofstede's (2001) conceptualization of power

4 We did not check the manipulation of countercultural priming because explicitlyasking people about psychological states pertaining to high or low power distance dur-ing the experiment tends to undercut the manipulation. That is, asking people for theirbeliefs about low or high power distance during the experiment may interfere with theeffects that the priming manipulation is supposed to have (namely to make counter-cultural values regarding power distance a salient issue for one group of participantsand not for the other group of participants). Furthermore, from earlier studies (e.g.,Greenberg et al., 1997; Loseman et al., 2009; Van den Bos, 2001; Van den Bos &Miedema, 2000; Van Prooijen et al., 2002) we know that asking participants to com-plete questions that ask them to think about certain issues indeed make these issuesmore salient than not asking them to complete questions about these issues.

5 As is common with these kinds of salience manipulations (see, e.g., Greenberg etal., 1997; Loseman et al., 2009; Van den Bos, 2001; Van den Bos & Miedema, 2000;Van Prooijen et al., 2002), all participants completed the Positive and Negative AffectSchedule (PANAS; Watson, Clark, & Tellegen, 1988) as a filler task between the coun-tercultural priming manipulation and the voice versus no-voice manipulation. ThePANAS consists of two ten-item subsets, one measuring positive affective states (PA;alpha=.91) and one measuring negative affective states (NA; alpha=.86). Controllingfor positive and negative affective states still yielded the predicted three-way interac-tion between country, countercultural priming, and procedure on treatment satisfac-tion, F(1, 240)=8.81, pb .01, ηp

2=.04. Thus, affective states cannot explain the effectsreported and therefore PA and NA were dropped from the analyses.

distance, in the experimental condition in which high power distancewas primed among our Dutch participants instructions were asfollows:

This part of the study will focus on other people's potential to de-termine or direct your behavior. More specifically, we ask you toread some materials and answer some questions that ask you toimagine that you are the less powerful and are willing to acceptthat those who have power over you (e.g., employers, politicians,police officers, etc.) have this because of their formal, hierarchicalposition.

Question 1: Please describe a situation out of your own life inwhich there was a large distance between you and the personwho formally had power over you. Thus, we ask you to imagineand describe to us a situation in which you were willing to acceptthat a person had power over you because of the person's formal,hierarchical position. Could you briefly describe this situation tous?Question 2: Imagine there is a large distance between you and aperson who has power over you, thus that you are willing to ac-cept that a person who formally has power over you because ofthe person's formal, hierarchical position. Could you briefly de-scribe how you would feel in such a situation and why it may bea good thing when a person with power occupies this powerfulposition by means of a formal, hierarchical appointment?

Similar to the control conditions in earlier salience studies, par-ticipants randomly assigned to the control condition in whichpower distance was not primed completed two filler questions thathad been shown in earlier studies to be neutral and not to instigateaffective reactions among participants (e.g., Greenberg et al., 1997;Loseman, Miedema, Van den Bos, & Vermunt, 2009; Van den Bos,2001; Van den Bos & Miedema, 2000; Van Prooijen et al., 2002). Spe-cifically, participants in the control conditions received the followinginstructions:

In this part of the study we will ask you to read some materialsand answer some questions concerning watching television. Weask you that you carefully read them and complete them.

Question 1: Please describe briefly the thoughts and emotions thatcome to mind when you think of the concept watching TV.Question 2: Please describe a situation out of your own life inwhich watching TV played a role.4, 5

Then, in an ostensibly unrelated part of the study, the proceduremanipulation was induced. This manipulation was similar to the pro-
cedure manipulation used by Brockner et al. (2001, Studies 1 and 2).All participants were asked to read and describe their reactions to thefollowing situation:
Imagine that you are working in a company that has just undergonemajor changes. It is now usingmore advanced technology to manu-facture and sell different products in new areas. As a result, your de-partment also changed dramatically. It is smaller now, with feweremployees and managers. There are new reporting structures quitedifferent fromwhat you have been used to. Your ownwork has beeninfluenced directly. What you do each day is very different fromwhat you used to do and in fact continues to change as the companymoves forward. Some employees are excited about the changes be-ingmade. Others are not as satisfied. Overall, it is still too early to tellwhether these changeswill ultimately place the company in a betterposition than it was before the changes started. As a result of thechanges you have a new manager in your department.

In the voice condition, the story continued with the following text:

As decisions about the changes are now being made, the new man-ager is happy to consult with you and others aboutwhat changes arebeing considered and whether you think they are good ideas. Thismanager also holds a weekly meeting in which you and your col-leagues can suggest any ideas you and your colleague's thought ofin the past week. After listening to your ideas and considering otheropinions, this manager makes the final decision.

In the no-voice condition, participants read the following text:

As decisions about the changes are now being made, the newmanager does not consult with you or others about what changesare being considered and whether you think they are good ideas.Attempts to consult with this manager about what changes are be-ing considered are ignored. This manager makes the decisions byhimself and then announces them to you.

In the final part of the study, participants completed questionsthat assessed their reactions to the scenario. As a check on the

1

2

3

4

5

6

7

Absent Present Absent Present

Sat

isfa

ctio

n

Countercultural Priming Countercultural Priming

INDIA NETHERLANDS

Voice

No Voice

Fig. 1. Satisfaction with treatment among Indian and Dutch participants as a function of voice versus no-voice procedures and the presence versus absence of the priming of thecountercultural norm regarding power distance. Error bars represent standard errors of the mean.

6 Inspecting Cook's (1977) distance measure in the main analyses within our Indianand Dutch samples (Cohen et al., 2003) revealed that 3 participants (1 in the Indiansample and 2 in the Dutch sample) had Cook's distance scores more than 3 SDs abovethe mean. These participants were excluded from all analyses reported. Please notethat although the answers of at least some of these participants may have deviatedfrom what was expected, not excluding the 3 participants continued to yield the pre-dicted three-way interaction between country, countercultural priming, and procedureon participants' ratings of treatment satisfaction, F(1, 245)=6.96, pb .01, ηp

2=.03,attesting to the robustness of the effect predicted.


procedure manipulation, participants rated their level of agreementwith the following statements: “The newmanager in this situation at-tempts to involve me in decisions that affect me,” and “My views aretaken into account by the manager.” These items were modeled afterBrockner et al. (2001, Studies 1 and 2) and Tyler (1987). Responsescould range from not at all (1) to very much (7). Participants' answersto the questions were averaged to form a reliable check of the proce-dure manipulation (alpha=.79).

The dependent variable asked participants how pleased, disap-pointed, enthusiastic, and discouraged they felt about the way theywere treated by their manager (1=not at all, 7=very much). Thesequestions were assessed among filler items and after reverse scoringthe answers to the second and fourth questions, participants' answersto these four items were averaged to form a reliable scale of treat-ment satisfaction (alpha=.86), such that higher scores reflectedgreater satisfaction. Moreover, coefficient alphas were high and virtu-ally identical within the Netherlands and India (.89 and .86, respec-tively), supporting the notion that the dependent variable attainedmeasurement equivalence in the two countries. At the end of thestudy, participants were thoroughly debriefed.

Results

Manipulation check

A 2 (country: India vs. Netherlands)×2 (countercultural priming:absent vs. present)×2 (procedure: voice vs. no voice) analysis of var-iance on the scale that checked the procedure manipulation yieldedtwo significant effects. First, there was a very strong main effect ofthe procedure manipulation, F(1, 242)=307.00, pb .001, ηp

2=.56. Asintended, participants thought their views were taken more into con-sideration in the condition inwhich theywere allowed an opportunityto voice their opinions (M=5.41, SE=0.09) than in the conditionin which they were not allowed such an opportunity (M=2.82,SE=0.14). There also was a significant two-way interaction betweencountry and procedure, F(1, 242)=27.48, pb .001, ηp

2=.10, whichshowed that whereas the procedure effect was highly significant inboth countries, it was greater in the Netherlands, F(1, 242)=268.97,pb .001, ηp

2=.53, than in India, F(1, 242)=77.03, pb .001, ηp2=.24.

Treatment satisfaction

A 2 (country)×2 (countercultural priming)×2 (procedure) analysisof variance onparticipants' ratings of treatment satisfaction yielded foursignificant effects: a main effect of procedure, F(1, 242)=207.37,pb .001, ηp

2=.46, a main effect of country, F(1, 242)=71.27, pb .001,

ηp2=.23, a two-way interaction between country and procedure, F(1,

242)=7.10, pb .01, ηp2=.03, and the predicted three-way interaction

between country, countercultural priming, and procedure, F(1,242)=9.95, pb .01, ηp

2=.04. Fig. 1 shows the effects together with therespective error bars.6

The main effect of the procedure manipulation showed that par-ticipants were more satisfied with the way they were treated whenthey received an opportunity to voice their opinions than whenthey did not receive such an opportunity. The main effect of countryindicated that our Indian participants showed higher satisfaction rat-ings than our Dutch participants. The country×procedure effectshowed that the procedure effect was stronger in the Netherlandsthan in India, a finding which is in accordance with people reactingmore strongly to voice versus no-voice procedures in low powerdistance cultures as opposed to high power distance cultures. Theseeffects were significantly qualified by the predicted three-way inter-action between country, countercultural priming, and procedure.

To specify the nature of the three-way interaction we conductedwithin-culture analyses to evaluate the interaction effect betweenprocedure and countercultural priming within each culture. Theseanalyses indeed showed that the countercultural priming×procedureinteraction was significant within each culture and, as predicted byour hypothesis, that these interaction effects took different forms inthe two cultures.

Indian resultsA 2 (countercultural priming)×2 (procedure) analysis of variance

on the satisfaction ratings of our Indian participants yielded a maineffect of procedure, F(1, 117)=52.10, pb .001, ηp

2=.31, and the pre-dicted two-way interaction effect between countercultural primingand procedure, F(1, 117)=4.41, pb .04, ηp

2=.04. In support of ourpredictions we found in the control condition in India, in which lowpower distance was not primed, that participants were more satisfiedwith the way they had been treated by their supervisor when they re-ceived voice (M=4.99, SE=0.23) than when they were denied voice(M=3.87, SE=0.23), F(1, 117)=13.24, pb .001, ηp

2=.10. However,when low power distance was primed among the Indian participantsthey responded much more strongly to receiving voice (M=5.31,


SE=0.20) versus not receiving voice (M=3.27, SE=0.20), F(1,117)=46.18, pb .001, ηp

2=.28. In fact, the effect of the procedure ma-nipulation on treatment satisfaction was almost three times strongeramong the Indian participants when low power distance was primedthan when it was not primed (ηp

2=.28 and ηp2=.10, respectively).

Thus, as hypothesized, the procedure effect on treatment satisfactionwas significantly stronger among our Indian participants when thepossibility of having a small distance with power distance was primedthan when it was not primed.

It can also be noted that the countercultural priming manipulationhad a significant effect on treatment satisfaction of our Indian partici-pants in the no-voice condition, F(1, 117)=6.95, pb .02, ηp

2=.06, suchthat they were less satisfied when low power distance was primedrather thanwhen itwas not primed. The simple effect of the countercul-tural priming manipulation was not statistically significant among ourIndian participants in the voice condition, F(1, 117)=0.64, p>.42,ηp2=.01. We will return to these findings in the Discussion.

Dutch resultsA 2 (countercultural priming)×2 (procedure) analysis of variance

on the satisfaction ratings of ourDutchparticipants yielded amain effectof the procedure manipulation, F(1, 125)=203.45, pb .001, ηp

2=.62,and the predicted two-way interaction effect between counterculturalpriming and procedure, F(1, 125)=5.85, pb .02, ηp

2=.04. In furthersupport of our predictions we found in the control condition in theNetherlands, in which high power distance was not primed, that partic-ipants were much more satisfied with the way they had been treatedwhen they received voice (M=4.66, SE=0.18) than when they werenot allowed voice (M=1.97, SE=0.09), F(1, 125)=145.06, pb .001,ηp2=.54. When high power distance was primed among our Dutch par-

ticipants they responded in a significant but weaker way to receivingvoice (M=4.08, SE=0.23) versus not receiving voice (M=2.17, SE=0.11), F(1, 125)=67.53, pb .001, ηp

2=.35. In fact, the effect of the proce-dure manipulation on treatment satisfaction was more than 1.5 timesweaker among our Dutch participants when low power distance wasprimed (ηp

2=.35) than when it was not primed (ηp2=.54). Thus, as hy-

pothesized, the procedure effect on treatment satisfaction was signifi-cantly weaker among our Dutch participants in the condition in whichthe possibility of having a large distancewith power holders was (as op-posed to was not) primed.

It can also be noted that the countercultural priming manipulationhad a significant effect on treatment satisfaction of our Dutch partici-pants in the voice condition, F(1, 125)=9.13, pb .01, ηp

2=.07, suchthat these participants felt more satisfied when high power distancewas not primed than when it was primed. The effect of the countercul-tural priming manipulation was not statistically significant among ourDutch participants in the no-voice condition, F(1, 125)=0.64, p>.23,ηp2=.01. We will come back to these findings in the Discussion.

Discussion

This study delineates important conditions under which the effectof being allowed versus denied voice on experienced treatment satis-faction will be more versus less likely to emerge. Our findings showthat when distance to power holders is large, either because of na-tional culture (which was likely to be the case in the control conditionamong the sample from India) or because of situational cues that re-mind people that power distance can be large (as was the case in theexperimental condition among the sample from the Netherlands), thevoice effect is considerably weaker than when power distance is smalleither because of national culture (as in the control condition amongthe sample from the Netherlands) or situational cues (as in the exper-imental condition among the sample from India). Furthermore, thefact that the independent variables of countercultural priming andprocedure were experimentally manipulated allows us to draw causalinferences with a high degree of confidence.

Moreover, significant effects of power distance were obtained as aresult of relatively subtle inductions of countercultural values thatconsisted of participants completing two open-ended questions only.In other words, people's reactions to voice deviated from presumablydeeply-ingrained cultural beliefs in power distance merely by askingthem to contemplate a countercultural level of power distance, thatis, low power distance in India and high power distance in theNetherlands. This is one indication of the strength and utility of thecountercultural–experimental approach to the study of cross-culturaldifferences advocated in the current paper.

In an earlier paper in which we explored the relevance of the coun-tercultural experimental approach to differences between masculineand feminine cultures (Van den Bos et al., 2010), the effects of voiceand countercultural priming were moderated by differences in partici-pants' capabilities to perform the task at hand and hence their expertiseof providing relevant input (cf. Van den Bos & Spruijt, 2002). Such amoderator was not needed in the current paper to reveal significant in-teraction effects pertaining to relevant cross-cultural differences. Thisattests to the robustness of the countercultural experimental approachto cross-cultural effects advocated. This also suggests that sometimesincluding moderating variables in experimental cross-counterculturaldesigns is needed to obtain relevant effects pertaining to cross-cultural differences (cf. Van den Bos et al., 2010), and that sometimesthesemoderators are not needed and that hence simpler designs can re-veal relevant cross-cultural effects (cf. the current paper). Future re-searchers should be aware of the inclusion of possibly relevantmoderators in their cross-cultural studies when using the experimentalcountercultural priming approach examined here and in the Van denBos et al. (2010) paper.

Implications

The present findings have implications for cross-cultural researchin at least two ways. First, they enable us to show that cultural differ-ences in people's responsivity to voice and no-voice procedures aredue to power distance with a research design entailing greater inter-nal validity than those employed in previous research. In doing so, thepresent study revealed that cultural differences in power distance andhow this affect people's reactions to the presence and absence ofvoice can be moderated in predictable and reliable ways by partici-pants being primed with countercultural values pertaining to powerdistance. This is a new finding, not earlier reported in the research lit-erature on voice and power distance (e.g., Brockner et al., 2001) orprocedural justice and power distance (e.g., Lam et al., 2002; Lee etal., 2000; Tyler et al., 2000; Yang et al., 2007; Zhang & Begley, 2011).

Second, the method of manipulating countercultural psychologicalstates offers cross-cultural researchers an important methodologicaltool that may be used more broadly to examine the causal impact ofa wide array of cultural influences on people's beliefs and behaviors.A common research strategy employed by cross-cultural researchersis to select participants from different cultures and assume that theparticipants differ along certain psychological dimensions, which, inturn, elicit differences in dependent variables of cognition, emotion,or behavior. An exemplar of this practice would be if we only consid-ered the results obtained in the control condition of the presentstudy. As mentioned earlier (e.g., Hofstede, 2001), and consistentwith the pilot test results presented in Footnote 1, people from theNetherlands can be assumed to be lower in power distance thanthose from India. Therefore, if we were to examine only the partici-pants in the control condition in their responsivity to voice andno-voice procedures, we would expect to find an interaction effectbetween country and procedure, such that the difference betweenthe voice and no-voice conditions should be more pronounced inthe Netherlands than in India. Indeed, upon conducting such a 2(country)×2 (procedure) analysis of variance in the control condi-tion only, we found a significant interaction effect between country


and procedure, F(1, 121)=17.24, pb .001, ηp2=.12, which showed

that the effect of the procedure manipulation on treatment satisfac-tion was about three times stronger in the Netherlands (ηp

2=.47)than in India (ηp

2=.15).Itwould be tempting to conclude from the results in the control con-

dition only that the country difference in response to the procedurema-nipulation was attributable to power distance beliefs. However, itwould not be possible to draw such an inference with a high degree ofconfidence because any of a host of differences between the partici-pants from the Netherlands and those from India could have produceddifferences in their responsivity to voice versus no-voice procedures inthe control condition. One way to address this issue would be to mea-sure various psychological dimensions that may have accounted forthe cross-cultural differences and then statistically control for these di-mensions in subsequentmediational analyses. This noted, for reasons ofinternal validity the experimental method used in the present researchhas important advantages. By randomly assigning half of the partici-pants in India and in the Netherlands to a countercultural condition(in which low power distance was primed in India and high powerdistance was primed in the Netherlands), we may gain greater confi-dence that country or cultural differences in power distance accountedfor the results in the control condition. That is, priming those from Indiawith low power distance should heighten the voice effect whereaspriming those from the Netherlands with high power distance shouldreduce the voice effect. In fact, this is precisely what the present resultsshowed.

The study by Van den Bos et al. (2010) indicated that cultural dif-ferences in masculinity and femininity dictated whether the voiceeffect on satisfaction was exhibited to a greater extent by those whosaw themselves as more capable versus less capable of providingmeaningful input into a decision process. Relatedly, Bechtoldt, DeDreu, Nijstad, and Choi (2010) revealed that priming Dutch partici-pants with a countercultural norm of appropriateness induced themto be more appropriate in a brainstorming task when participantswere focused on the task. Building on and extending the currentstudy as well as studies earlier reported in the research literature(e.g., Bechtoldt et al., 2010; Oyserman, 2011; Van den Bos et al.,2010), future studies may draw on the experimental method usedin the present research to examine causal relationships betweenother cultural dimensions (such as individualism and collectivism)and what people think, feel, and do.

Furthermore, the countercultural priming method does more thanoffer a more internally valid way of examining cross-cultural differ-ences than do procedures in which the independent variables aremeasured rather than manipulated. It also may help to reduce prob-lems associated with the most typical way of measuring the indepen-dent variables, which is via self-report. Cultural differences in socialdesirability and response styles may cloud interpretation of resultsin which the independent variables are measured through self-report. Ensuring measurement equivalence (i.e., verifying that theitems have the same meaning to people from different cultures)also may be a challenge. We note explicitly that we are not proposingthat the current countercultural priming method is a panacea for allthe problems researchers encounter when studying cross-culturaldifferences, but we do propose that by manipulating culturally-relevant independent variables with the method used in the presentstudy researchers may be able to reduce a variety of methodologicalconcerns associated with measuring independent or intervening vari-ables in cross-cultural research.

Limitations and future directions

Whereas the present research yielded support for our hypotheses,it has both empirical and methodological limitations which, in turn,point to important areas for future research. At the empirical level,the effects produced by the experimental manipulations of

countercultural priming and procedure were consistent across thetwo cultures in showing that the countercultural prime altered thetypical relationship between the procedure manipulation and satis-faction. However, the results were not entirely consistent across thetwo cultures, in the following sense: within India there was a signifi-cant effect of the countercultural prime within the no-voice condition(such that participants were less satisfied when low power distancewas primed than they were in the control condition), but there wasno effect of the countercultural prime within the voice condition. Incontrast, within the Netherlands there was a significant effect of thecountercultural prime within the voice condition (such that partici-pants were less satisfied when high power distance was primedthan they were in the control condition), but there was no effect ofthe countercultural prime within the no-voice condition.

Howmightwe explain the different effects of the primingmanipula-tion in the two cultures within the voice and no-voice conditions? Onepossibility is that negative expectancy violations are more upsettingthan positive expectancy violations are satisfying, or as Kahneman andTversky (1984) put it, “losses loom larger than gains.” In India, the coun-tercultural prime put participants in a low power distance frame ofmind, in which they may have seen themselves as being more entitledto have voice. When they were not given voice, however, they mayhave felt negatively violated and dissatisfied as a result. In the Nether-lands, the countercultural prime put participants in a high power dis-tance mindset, in which they may have seen themselves as being lessentitled to have voice. When they were then given voice, they mayhave been somewhat pleasantly surprised, but not to the point thatthey would feel more satisfied than they did in the control condition.

Conceptually similar results emerged in the control condition, inwhich the magnitude of the culture effect on satisfaction differed inthe no-voice and voice conditions. That is, in the control conditionparticipants from the low power distance culture were significantlyless satisfied with the absence of voice than were those from thehigh power distance culture, F(1, 242)=59.07, pb .001, ηp

2=.20.However, satisfaction with the presence of voice did not differ be-tween the low and high power distance cultures, F(1, 242)=1.33,p>.24, ηp

2=.01. Perhaps the notion of losses looming larger thangains may account for these findings as well (Price et al., 2001) suchthat when those with low power distance were negatively violatedby not having voice they were particularly dissatisfied. However,when those with high power distance were presumably pleasantlysurprised by having voice, they did not show an especially highlevel of satisfaction. In any event, future research is needed to betterunderstand the asymmetric results that emerged within the voiceand no voice conditions as a function of the countercultural primingmanipulation and participants' culture.

In showing that power distance moderated participants' reactionsto voice, we are not suggesting that power distance is the onlyculturally-relevant variable that will moderate people's reactions tothe presence or absence of voice, nor are we implying that power dis-tance is the only relevant variable that distinguishes the Indian andDutch cultures from each other. Of course, it is possible that otherelements than power distance have affected some components ofour participants' reactions. This is always possible. And, of course,when conducting cross-cultural research not only power distance,but also individualism–collectivism and other cultural notions mayhave had an effect on the findings obtained, as they may have affectedthe results obtained in earlier studies on voice and power distance(Brockner et al., 2001) and procedural justice and power distance(Lam et al., 2002; Lee et al., 2000; Tyler et al., 2000; Yang et al.,2007; Zhang & Begley, 2011). This noted, consider the manipulationwe used to prime low power distance in India and high power dis-tance in the Netherlands. In this manipulation participants wereasked to imagine that they and those who have power over them re-gard each other as equals or that they are the less powerful and arewilling to accept that those who have power over you have this


because of their formal, hierarchical position. Thus, we think that wehave good reasons to argue that the wording used in our manipula-tion of power distance captured the essence of power distance as for-mulated by Hofstede (2001) and others (e.g., Heine, 2010; Kirkmanet al., 2009; Kirkman & Shapiro, 2001).

We deliberately constructed the cross-cultural priming manipula-tion to capture the dimension of power distance in particular. In doingso, we did our utmost best to ensure that this manipulation would notbe related to individualism–collectivism or other cultural dimensionsidentified in the literature (see, e.g., Au, Hui, & Leung, 2001; Chiu &Hong, 2007; Cohen & Leung, 2009; Heine, 2010; Hofstede, 2001; Hui& Triandis, 1985; Leung, 2005; Markus & Kitayama, 1991). Again,inspecting the materials used in our manipulation of power distanceand the definition of power distance as formulated by Hofstede(2001) and others (e.g., Heine, 2010; Kirkman et al., 2009; Kirkman &Shapiro, 2001) suggests that this manipulation was in all likelihood in-dependent of constructs such as individualistic versus collectivistic cul-tures and was related to power distance directly. In other words, inconstructing the countercultural priming manipulation, great care wastaken to make the manipulation unrelated to the other dimensionsthat can be reflective of ways in which people from India and theNetherlands differ, which could also account for the present findings.But, again, future research with different stimulus materials and differ-ent samples from different cultures is needed to examine in detail therobustness of the experimental method and specific stimulus materialsused here. Thus,we are stating explicitly that future research on the dis-criminant validity of our method of priming countercultural power dis-tance beliefs is warranted. We hope that the current findings, togetherwith earlier reported studies (e.g., Bechtoldt et al., 2010; Oyserman &Lee, 2008; Van den Bos et al., 2010), will stimulate cross-cultural re-search to use the countercultural experimental approach to exploreother important issues in the psychology of culture and cross-culturaldifferences.

We also suggest that future cross-cultural research may profit froman explicit consideration of multi-dimensional comparisons betweendifferent countries studied. That is, most cross-cultural studies focuson one important cultural dimension. The current study is no exception.However, we propose that the psychology of cross-cultural differencesmay be enhanced considerably if studies were to include multiple cul-tural dimensions simultaneously. For example, it may be useful to ma-nipulate multiple countercultural primes orthogonally of one another(e.g., power distance×individualism–collectivism).

At a methodological level, whereas the experimental design in thecurrent study allowed for a high degree of internal validity, future re-search should evaluate the external validity of the present findings.For instance, we examined the moderating impact of power distanceon the relationship between voice and treatment satisfaction in a con-text in which participants indicated how they would respond to a sit-uation they were asked to imagine, which may be different from howthey would respond to a situation that they actually experienced. It isreassuring that previous research has shown that cultural differencesin measured power distance similarly moderated people's reactionsto voice in a context in which people actually experienced the events,rather than being asked to imagine how they would respond (e.g.,Brockner et al., 2001). Relatedly, the other study using the counter-cultural priming technique with respect to voice (Van den Bos etal., 2010) examined people's satisfaction with a decision that theyactually experienced rather than a hypothetical one. Similarly,Bechtoldt et al., 2010) used actual brainstorming tasks when examin-ing how countercultural norms influenced creativity. In other words,it is unlikely that the present findings are an artifact of or limited topeople's predictions of how they believe they would respond, ratherthan how they would actually respond. Nevertheless, future researchis needed which examines the impact of priming counterculturalpower distance beliefs on people's reactions to the presence or ab-sence of voice in the context of events that they actually experience.

Future research also should show the effects we reported here usingdifferent operationalizations of procedural fairness, different salienceor priming manipulations, different dependent variables, and differ-ent samples from different high and low power distance countries.We hope that the study reported here, combined with other develop-ments (see, e.g., Oyserman; Oyserman & Lee, 2008), will stimulatesuch future research endeavors.

A careful reader may wonder why we did not also include a “pro-cultural priming” condition in our design (i.e., a high-power-distanceprime in the Indian sample and a low-power-distance prime in theDutch sample). In response to this interesting issue we note that inthe case of cultural norms people often do not realize to whatnorms they adhere exactly. That is, many of our cultural norms arenot salient and we tend to treat them as default. This may not onlyhamper the accurate reporting of cultural norms (an issue to whichwe alluded earlier), but this also suggests that not reminding partici-pants about their cultural norms may be the most natural controlcondition in experimental cross-cultural research. After all, peopletend to go through life without their cultural norms being made sa-lient all the time. Thus, not making cultural norms salient may bethe most logical control condition.

Furthermore, we are not certain what to expect when pro-culturalnorms are made salient: When people are reminded about what theyassume to be default in their culture they may show the same effectas they normally do (and hence we would not expect differentvoice effects in pro-cultural priming versus nothing-primed condi-tions). However, it also possible that unequivocally priming withpro-cultural values (i.e., high power distance in India and lowpower distance in the Netherlands) would lead to different voice ef-fects (i.e., weaker voice effects in India and stronger voice effects inthe Netherlands) compared to nothing-primed conditions. Moreover,there is the possibility that when pro-cultural norms are made salientthat people start to think and deliberate about their cultural norms,more so than when they are not reminded explicitly about their cul-tural norms. Compared to a condition in which they are not remindedabout their cultural norms, this may cause them to switch from an ex-periential mindset to a more rationalistic mindset and hence to reactdifferently toward voice and no-voice procedures (see Maas & Vanden Bos, 2009; Van den Bos & Maas, 2012).

In short, we think that by comparing experimental conditions inwhich countercultural norms were primed with control conditions inwhich participants completed two questions that were shown in earlierstudies to be neutral and not to trigger strong affective reactions (e.g.,Greenberg et al., 1997; Loseman et al., 2009; Van den Bos, 2001; Vanden Bos & Miedema, 2000; Van Prooijen et al., 2002), we focused onthe most relevant contrast for the current purposes of being the firstto use a countercultural–experimental approach to the study of powerdistance and voice. This noted, we would applaud if different controlconditions would be used in future research.

In the onepaper inwhichwe explored our experimental countercul-tural approach to cross-cultural differences we focused on “masculine”and “feminine” values (Van den Bos et al., 2010). Participants in thecountercultural conditions in that paper were asked to assign impor-tance to either masculine or feminine values, depending on whetherparticipants were frommore “feminine” cultures (i.e., the Netherlands)or more “masculine” cultures (i.e., the USA; Hofstede, 2001). Whereasthe results in those countercultural conditions might have beencompromised by demand characteristics, the results of the current ex-periment are less likely to be vulnerable to that issue. That is, in the cur-rent study participants took part in several parts that ostensibly wereunrelated to each other. Furthermore, in the countercultural (experi-mental) conditions participants merely were asked to complete onlytwo open-ended questions that asked them to think of their thoughtsand emotions of having either a low or high distance to power holders.It is unlikely that this may have forced participants to respond in differ-ent ways to the scenarios we then presented to them. Furthermore, we


think this manipulation was subtler than previous manipulations usedin the power distance literature (see, e.g., Brockner et al., 2001,Study 4). Moreover, even if participants saw the tasks as related toone another, the primary prediction was reflected in a fairly complexthree-way interaction effect. Though not impossible, it seems unlikelythat the emergence of the three-way interaction was the result of par-ticipants discerning our prediction and wanting to act in a manner con-sistent with it, especially given that all of our independent variablesconsisted of between-subjects factors. All this said, we hasten to notethat future research with other research methods, more impactfulreactions, and including not only satisfaction ratings but also justicejudgments and behavioral effects, would be important. This may con-tribute to insights into the voice effect in different cultures and mayhelp to further delineate the proposed mechanism to study cross-cultural differences experimentally.

The findings obtained in the experimental conditions also suggestthe relative ease with which people may be situationally induced tobehave in ways contrary to the prevailing values in their cultures.Specifically, we found that when low power distance was made sa-lient among Indian participants, voice versus no-voice opportunitieshad a much larger impact on their treatment satisfaction, whereaswhen high power distance was made salient among participantsfrom the Netherlands, the voice effect on satisfaction was consider-ably reduced. In other words, the effects produced by years of culturalupbringing may be dramatically altered (at least in the short-term) asa result of exposure to relatively subtle inductions such as the com-pletion of two open-ended questions.

Having said that, we would not characterize the present partici-pants as completely indifferent to their extant power distance beliefs.The results showed that in the control condition, in which partici-pants' extant power distance beliefs were unconstrained, the proce-dure effect was a lot stronger in the Netherlands (the lower powerdistance culture) than in India (the higher power distance culture).If participants were largely indifferent to their extant power distancebeliefs, then the procedure effect should have been reversed in thecountercultural priming condition: those primed to be low in powerdistance (the Indian sample) should have shown a stronger proce-dure effect than those primed to be high in power distance (theDutch sample). This did not happen, however. The magnitude of theprocedure effect in the countercultural priming condition in Indiawas virtually identical to the procedure effect in the counterculturalpriming condition in the Netherlands, reflected in a nonsignificantculture×procedure interaction within the countercultural primingcondition, F(1, 121)=0.12, p>.73, ηp

2=.00. Thus, as predicted, thecountercultural priming manipulation significantly reduced (indeed,neutralized) the effect of participants' extant power distance beliefson their satisfaction. The priming manipulation did not, however,completely overturn the influence of participants' extant beliefs.

Finally, another important matter of external validity to beaddressed in future research pertains to the generality of the presentfindings. Perhaps the countercultural priming manipulation workedparticularly well on the present participants. As adolescents in theirearly 20s, they may not be deeply committed to power distance be-liefs proscribed by their cultures. Whereas the pilot study suggestedthat participants from India likely had higher power distance beliefsthan their counterparts from the Netherlands, future research shouldexamine people's extant level of power distance beliefs as well astheir degree of commitment to such beliefs, independent of thelevel of power distance. It may be that the present results will beless likely to emerge among people who are highly committed toculturally-proscribed power distance beliefs. In other words, thecountercultural priming manipulation may be less likely to “take”among individuals who are highly committed to their extant powerdistance beliefs.

Related to this, the countercultural primingmethodology is likely tobe more effective to the extent that participants are “primable” (that is,

to the extent that the construct can be easily manipulated). Further-more, individual difference variables may affect how much peoplemay be primable by culturally-varying dimensions. Chiu, Morris,Hong, and Menon (2000) showed that people high in need for closure(Kruglanski &Webster, 1996) are more likely to buy into the prevailingbelief system in their culture. We would therefore surmise that thecountercultural priming methodology may be less effective amongpeople high in need for closure. Other individual difference variablesalso may be related to people's susceptibility to being primed, such asthe Big Five dimension of openness to experience (Costa & McCrae,1985). Furthermore, there may even be cultural differences in people'sprimability. Consider the dimension of uncertainty avoidance, whichdeals with a society's tolerance for uncertainty and ambiguity(Hofstede, 2001). Uncertainty avoidance refers to the extent to whicha culture induces its members to feel either uncomfortable in situationsthat are novel, unknown, surprising, and different from usual, in aword,unstructured. Perhaps people will be less primable in high uncertaintyavoidance cultures.

To conclude, factors affecting people's primability speak to the limi-tations of the countercultural priming approach used in the presentstudy. Whenever people are more adherent to the prevailing beliefs/values in their culture, the priming method we are advocating may beless effective. That said, we are buoyed by the results of the presentand previous study (Van den Bos et al., 2010) suggesting that the coun-tercultural priming method offers researchers a useful way to test forcultural differences in people's beliefs and behaviors with a high degreeof internal validity. Thus, we think that the current paper provides animportant step to move from correlational to more experimental ap-proaches to the study of cross-cultural differences.

References

Ang, S., Van Dyne, L., & Begley, T. M. (2003). The employment relationship of foreignworkers versus local employees: A field study of organizational justice, job satis-faction, performance, and OCB. Journal of Organizational Behavior, 24, 561–583.

Au, K., Hui, M. K., & Leung, K. (2001). Who should be responsible? Effects of voice andcompensation on responsibility attribution, perceived justice, and post-complaintbehaviors across cultures. International Journal of Conflict Management, 12,350–364.

Baumeister, R. F. (2005). The cultural animal: Human nature, meaning, and social life.New York: Oxford University Press.

Bechtoldt, M. N., De Dreu, C. K. W., Nijstad, B. A., & Choi, H. S. (2010). Motivated infor-mation processing, social tuning, and group creativity. Journal of Personality andSocial Psychology, 99, 622–637.

Brockner, J. (2003). Unpacking country effects: On the need to operationalize the psy-chological determinants of cross-national differences. In R. M. Kramer, & B. M. Staw(Eds.), Research in organizational behavior, Vol. 25. (pp. 335–369). Greenwich, CT:JAI Press.

Brockner, J., Ackerman, G., Greenberg, J., Gelfand, M. J., Fransesco, A. M., Chen, Z. X.,et al. (2001). Culture and procedural justice: The moderating influence of powerdistance on reactions to voice. Journal of Experimental Social Psychology, 37,300–315.

Brockner, J., Heuer, L., Siegel, P. A., Wiesenfeld, B., Martin, C., Grover, S., et al. (1998).The moderating effect of self-esteem in reaction to voice: Converging evidencefrom five studies. Journal of Personality and Social Psychology, 75, 394–407.

Chiu, C. -Y., & Hong, Y. -Y. (2007). Cultural processes: Basic principles. In A. W.Kruglanski, & E. T. Higgins (Eds.), Social psychology: Handbook of basic principles(2nd ed.) (pp. 785–804). New York: Guilford Press.

Chiu, C. -Y., Morris, M. W., Hong, Y., & Menon, T. (2000). Motivated cultural cognition:The impact of implicit cultural theories on dispositional attribution varies as afunction of need for closure. Journal of Personality and Social Psychology, 78,247–259.

Cohen, D., & Leung, A. K. -Y. (2009). The hard embodiment of culture. European Journalof Social Psychology, 39, 1278–1289.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Appliedmultiple regression/correlationanalysis for the behavioral sciences (3rd ed.). Hillsdale, NJ: Erlbaum.

Cook, R. D. (1977). Detection of influential observations in linear regression.Technometrics, 19, 15–18.

Costa, P. T., Jr., & McCrae, R. R. (1985). The NEO personality inventory manual. Odessa, FL:Psychological Assessment Resources.

Fischer, R., & Smith, P. B. (2003). Reward allocation and culture: A meta-analysis. Journalof Cross-Cultural Psychology, 34, 251–268.

Fiske, S. T. (2006). Building bridges inside and outside social psychology: A case forlumping–neatly. In P. A. M. van Lange (Ed.), Bridging social psychology: Benefits oftransdisciplinary approaches (pp. 47–52). Mahwah, NJ: Erlbaum.


Folger, R. (1977). Distributive and procedural justice: Combined impact of “voice” andimprovement of experienced inequity. Journal of Personality and Social Psychology,35, 108–119.

Folger, R., Rosenfield, D., Grove, J., & Corkran, L. (1979). Effects of “voice” and peer opin-ions on responses to inequity. Journal of Personality and Social Psychology, 37,2253–2261.

Francesco, A. M., & Chen, Z. X. (2004). Collectivism in action: Its moderating effects onthe relationship between organizational commitment and employee performancein China. Group & Organization Management, 29, 425–441.

Greenberg, J., Solomon, S., & Pyszczynski, T. (1997). Terror management theory ofself-esteem and cultural worldviews: Empirical assessments and conceptualrefinements. In M. P. Zanna (Ed.), Advances in experimental social psychology,Vol. 29. (pp. 61–139). New York: Academic Press.

Heine, S. J. (2010). Cultural psychology. In S. T. Fiske, D. T. Gilbert, & G. Lindzey (Eds.),Handbook of social psychology (5th ed., Vol. 2, pp. 1423–1464). Hoboken, NJ: Wiley.

Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions,and organizations across nations (2nd ed.). Thousand Oaks, CA: Sage.

Hui, C. H., & Triandis, H. C. (1985). Measurements in cross-cultural psychology. Journalof Cross-Cultural Psychology, 16, 131–152.

Kahneman, D., & Tversky, A. (1984). Choices, values, and frames. The American Psychologist,39, 341–350.

Kirkman, B. L., Chen, G., Farh, J. -L., Chen, Z., & Lowe, K. (2009). Individual power distanceorientation and follower reactions to transformational leaders: A cross-level, cross-cultural examination. Academy of Management Journal, 52, 744–764.

Kirkman, B. L., & Shapiro, D. L. (2001). The impact of cultural values on job satisfactionand organisational commitment in self-managing work teams: The mediating roleof employee resistance. Academy of Management Journal, 44, 557–569.

Kruglanski, A. W., & Webster, D. M. (1996). Motivated closing of the mind: “Seizing”and “freezing”. Psychological Review, 103, 263–283.

Lam, S. S. K., Schaubroeck, J., & Aryee, S. (2002). Relationship between organizationaljustice and employee work outcomes: A cross-national study. Journal of Organiza-tional Behavior, 23, 1–18.

Lee, C., Pillutla, M., & Law, K. S. (2000). Power-distance, gender and organizationaljustice. Journal of Management, 26, 685–704.

Leung, K. (2005). How generalizable are justice effects across cultures? In J. Greenberg,& J. A. Colquitt (Eds.), Handbook of organizational justice: Fundamental questionsabout fairness in the workplace (pp. 555–586). Mahwah, NJ: Erlbaum.

Leung, K., Su, S., & Morris, M. W. (2001). When criticism is not constructive: Across-cultural investigation of responses to supervisory feedback as function ofinteractional justice. Human Relations, 54, 1155–1187.

Lind, E. A., Kanfer, R., & Earley, P. C. (1990). Voice, control, and procedural justice:Instrumental and noninstrumental concerns in fairness judgments. Journal ofPersonality and Social Psychology, 59, 952–959.

Lind, E. A., & Tyler, T. R. (1988). The social psychology of procedural justice. New York:Plenum.

Lind, E. A., Tyler, T. R., & Huo, Y. J. (1997). Procedural context and culture: Variation inthe antecedents of procedural justice judgments. Journal of Personality and SocialPsychology, 73, 767–780.

Loseman, A., Miedema, J., Van den Bos, K., & Vermunt, R. (2009). Exploring how peoplerespond to conflicts between self-interest and fairness: The influence of threats tothe self on affective reactions to advantageous inequity. Australian Journal ofPsychology, 61, 13–21.

Maas, M., & Van den Bos, K. (2009). An affective-experiential perspective on reactionsto fair and unfair events: Individual differences in affect intensity moderated byexperiential mindsets. Journal of Experimental Social Psychology, 45, 667–675.

Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition,emotion, and motivation. Psychological Review, 98, 224–253.

Martin, L. L. (1999). I-D compensation theory: Some implications of trying to satisfyimmediate-return needs in a delayed-return culture. Psychological Inquiry, 10,195–208.

Messick, D. M. (1993). Equality as a decision heuristic. In B. A. Mellers, & J. Baron (Eds.),Psychological perspectives on justice: Theory and applications (pp. 11–31). Cambridge,MA: Cambridge University Press.

Nisbett, R. E., & Ross, L. (1980). Human inference: Strategies and shortcomings of socialjudgment. Englewood Cliffs, NJ: Prentice-Hall.

Oyserman, D. (2011). Culture as situated cognition: Cultural mindsets, cultural fluency,and meaning making. In W. Stroebe, & M. Hewstone (Eds.), European Review ofSocial Psychology, Vol. 22. (pp. 164–214). Oxon, UK: Psychology Press.

Oyserman, D., & Lee, S. W. S. (2008). Does culture influence what and how we think? Ef-fects of priming individualism and collectivism. Psychological Bulletin, 134, 311–342.

Price, K., Hall, T., Van den Bos, K., Hunton, J., Lovett, S., & Tippett, M. (2001). Features ofthe value function for voice and their consistency across participants from fourcountries: Great Britain, Mexico, the Netherlands, and the United States. Organiza-tional Behavior and Human Decision Processes, 84, 95–121.

Schaller, M., & Crandall, C. S. (Eds.). (2004). The psychological foundation of culture.Mahwah, NJ: Erlbaum.

Spencer, S. J., Zanna, M. P., & Fong, G. T. (2005). Establishing a causal chain: Why exper-iments are often more effective than mediational analyses in examining psycho-logical processes. Journal of Personality and Social Psychology, 89, 845–851.

Thibaut, J., & Walker, L. (1975). Procedural justice: A psychological analysis. Hillsdale, NJ:Erlbaum.

Tyler, T. R. (1987). Conditions leading to value-expressive effects in judgments of proceduraljustice: A test of four models. Journal of Personality and Social Psychology, 52, 333–344.

Tyler, T. R., Lind, E. A., & Huo, Y. J. (2000). Cultural values and authority relations: Thepsychology of conflict resolution across cultures. Psychology, Public Policy, and Law,6, 1138–1163.

Tyler, T. R., Rasinski, K. A., & Spodick, N. (1985). Influence of voice on satisfaction withleaders: Exploring the meaning of process control. Journal of Personality and SocialPsychology, 48, 72–81.

Van den Bos, K. (1999). What are we talking about when we talk about no-voice pro-cedures? On the psychology of the fair outcome effect. Journal of Experimental So-cial Psychology, 35, 560–577.

Van den Bos, K. (2001). Uncertainty management: The influence of uncertaintysalience on reactions to perceived procedural fairness. Journal of Personality andSocial Psychology, 80, 931–941.

Van den Bos, K. (2003). On the subjective quality of social justice: The role of affect asinformation in the psychology of justice judgments. Journal of Personality and SocialPsychology, 85, 482–498.

Van den Bos, K. (2005). What is responsible for the fair process effect? In J. Greenberg,& J. A. Colquitt (Eds.), Handbook of organizational justice: Fundamental questionsabout fairness in the workplace (pp. 273–300). Mahwah, NJ: Erlbaum.

Van den Bos, K., Brockner, J., Stein, J. H., Steiner, D. D., Van Yperen, N. W., & Dekker, D. M.(2010). The psychology of voice and performance capabilities in masculine and fem-inine cultures and contexts. Journal of Personality and Social Psychology, 99, 638–648.

Van den Bos, K., & Maas, M. (2012). Adhering to consistency principles in an unjustworld: Implications for sense making, victim blaming, and justice judgments. InB. Gawronski, & F. Strack (Eds.), Cognitive consistency: A fundamental principle insocial cognition (pp. 424–444). New York: Guilford.

Van den Bos, K., & Miedema, J. (2000). Toward understanding why fairness matters:The influence of mortality salience on reactions to procedural fairness. Journal ofPersonality and Social Psychology, 79, 355–366.

Van den Bos, K., & Spruijt, N. (2002). Appropriateness of decisions as a moderator of thepsychology of voice. European Journal of Social Psychology, 32, 57–72.

Van den Bos, K., Van Lange, P. A. M., Lind, E. A., Venhoeven, L. A., Beudeker, D. A.,Cramwinckel, F. M., et al. (2011). On the benign qualities of behavioral disinhibition:Because of the prosocial nature of people, behavioral disinhibition can weaken plea-sure with getting more than you deserve. Journal of Personality and Social Psychology,101, 791–811.

Van den Bos, K., Vermunt, R., & Wilke, H. A. M. (1996). The consistency rule andthe voice effect: The influence of expectations on procedural fairness judgments.European Journal of Social Psychology, 26, 411–428.

Van Prooijen, J. -W., Van den Bos, K., & Wilke, H. A. M. (2002). Procedural justice andstatus: Status salience as antecedent of the fair process effect. Journal of Personalityand Social Psychology, 83, 1353–1361.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief mea-sures of positive and negative affect: The PANAS scales. Journal of Personality andSocial Psychology, 54, 1063–1070.

Yang, J. X., Mossholder, K. W., & Peng, T. K. (2007). Procedural justice climate and grouppower distance: An examination of cross-level interaction effects. The Journal ofApplied Psychology, 92, 681–692.

Yang, J. X., Mossholder, K. W., & Peng, T. K. (2009). Supervisory procedural justice effects:The mediating roles of cognitive and affective trust. The Leadership Quarterly, 20,143–154.

Zhang, Y., & Begley, T. M. (2011). Power distance and its moderating impact onempowerment and team participation. International Journal of Human ResourceManagement, 22, 3601–3617.

Documents

Delineating a method to study cross-cultural differences with experimental control: The voice effect and countercultural contexts regarding power distance