Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=vsoc20

Download by: [Rod Library, University of Northern Iowa] Date: 13 January 2016, At: 10:56

The Journal of Social Psychology

ISSN: 0022-4545 (Print) 1940-1183 (Online) Journal homepage: http://www.tandfonline.com/loi/vsoc20

Linguistic ostracism causes prejudice: Support fora serial mediation effect

Robert Thomas Hitlan, Michael A. Zárate, Kristine M. Kelly & M. CatherineDeSoto

To cite this article: Robert Thomas Hitlan, Michael A. Zárate, Kristine M. Kelly & M. CatherineDeSoto (2015): Linguistic ostracism causes prejudice: Support for a serial mediation effect, TheJournal of Social Psychology, DOI: 10.1080/00224545.2015.1119668

To link to this article: http://dx.doi.org/10.1080/00224545.2015.1119668

Accepted author version posted online: 17Nov 2015.Published online: 17 Nov 2015.

Submit your article to this journal

Article views: 14

View related articles

View Crossmark data

Linguistic ostracism causes prejudice: Support for a serialmediation effectRobert Thomas Hitlana, Michael A. Zárateb, Kristine M. Kellyc, and M. Catherine DeSotoa

aUniversity of Northern Iowa; bUniversity of Texas at El Paso; cWestern Illinois University

ABSTRACTThis research investigated the effects of linguistic ostracism, defined as anycommunication setting in which a target individual (or group) is ostracizedby another individual (or group) in a language that the target has extremelylimited ability to understand. Participants were included or ostracized bytheir group members during a computer-mediated group discussion. Halfof the ostracized participants were linguistically ostracized via their groupmembers conversing with one another in a language the participant didnot know well (Spanish Ostracism: SO), or in a language the participant didknow well (English Ostracism: EO). SO participants reported feeling lesssimilar than both included and EO participants. SO participants alsoreported being angrier and expressed more prejudice than included parti-cipants (and EO participants using effect size estimates). Results also pro-vided support for the hypothesized serial mediation model. Findings arediscussed in terms of implications for intergroup relations.

ARTICLE HISTORYReceived 16 September2013Accepted 5 November 2015

KEYWORDSExclusion; intergrouprelations; language;mediation; ostracism

Language use represents one of the primary routes through which people communicate with oneanother. As such, language use is closely interwoven into the cultural fabric of a given society. Theinability to effectively communicate with others at work or in a wide variety of social situations maybe stressful and have important implications for both intra- and intergroup relations. The currentresearch investigates how ostracism affects cognition and emotions and also examines some of thespecific downstream consequences associated with linguistic ostracism (i.e., prejudice). Generally,research in this area has tended to focus on specific outcomes associated with being ostracized (e.g.,psychological health, performance, conformity, aggression) or how these effects are moderated viaindividual differences (Williams, 2007). However, far fewer studies have asked the question “Whydoes ostracism affect targets?” To this end, we examine some of the specific mediational processesthat may serve to help in determining “why” ostracism elicits certain types of responses and, in doingso, help to fill an important gap in this area of research.

According to the US Census Bureau, since 1980 there has been over a 158% increase in thenumber of people who speak a language other than English at home (from 23 to 59.5 million; Ryan,2013). Of these, over half (37.5 million) indicated they spoke Spanish, with many reporting they hadlimited English proficiency. Further, approximately 23 million reported speaking a language otherthan English or Spanish (e.g., French, Russian, Persian, Chinese, Vietnamese) and, of these, approxi-mately 38% indicated that they had limited English proficiency. In addition, as of 2007, 4.5 millionindividuals indicated that they could not speak English at all. Given that language is an importantcomponent of ethnic identity (Cargile & Giles, 1997, 1998; Giles, Taylor, Lambert, & Albert, 1976;Giles, Williams, Mackie, & Rosselli, 1995) the continuing debate over multiculturalism within theUnited States is not surprising (Malkin, 2007; Zárate, Shaw, Marquez, & Biagas, 2012). In fact,

CONTACT Robert Thomas Hitlan rob.hitlan@uni.edu Department of Psychology, University of Northern Iowa, 437 BakerHall, Cedar Falls, IA 50614, USA.© 2015 Taylor & Francis

THE JOURNAL OF SOCIAL PSYCHOLOGYhttp://dx.doi.org/10.1080/00224545.2015.1119668

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

according to a Gallup poll, when asked about immigrants and language use, 96% of Americancitizens indicated that it is either important or essential for immigrants living in the United States tolearn English (US English, Inc., 2015). Here, it is assumed that this push for immigrants is less aboutmaking the immigrants feel comfortable and more about making the native English speakers feelcomfortable. Thus, one reason for the push toward an increased English language use/ability mayrelate to the general idea that people feel ostracized when others speak an unfamiliar languagearound them. This may explain why there are numerous instances of people being harassed,discriminated against, and even fired for speaking Spanish at their place of employment(Associated Press, 2000; Contreras, 2013; Fuchs, 2008; Garcia v. Gloor, 1980). Many such instancesinvolve individuals who know English but were simply speaking Spanish around others, whichpresumably made some people uncomfortable and/or violated an organization’s “English-only”rules. We argue that such instances may also serve to highlight group-based categorizations andassociated responses.

Research on social rejection, exclusion, and ostracism

The terms rejection, exclusion, and ostracism are often used interchangeably. However, it is possibleto distinguish between them in terms of correlates and outcomes associated with each (Bernstein &Claypool, 2012a, 2012b; Molden, Lucas, Gardner, Dean, & Knowles, 2009; Smart-Richman & Leary,2009; Williams, 2007). Williams (2007), defines ostracism as the ignoring of one or more individualsby one or more individuals, whereas exclusion involves keeping an individual (or group) apart fromothers, and rejection involves some sort of explicit statement that an individual (or group) isunwanted by others. The definition of ostracism noted above is most consistent with the behaviorsbeing assessed in the current research. However, for consistency, we will continue to use those termsused by previous researchers when describing their research below.

A large body of research has been amassed attesting to the aversive nature of exclusionarybehavior for one’s psychological and physical health (Baumeister & Leary, 1995; Williams, 2007).Generally speaking, such behavior can motivate targets to increase approach-oriented or conciliatorybehavior or, alternatively, increase avoidance-oriented or withdrawal forms of behavior. Increasedapproach-oriented behavior aims at reestablishing oneself as a good and worthy group member, suchas when one tries to compensate for a poorly performing group or one of its members (Williams &Sommer, 1997), or through increased conformity (Williams, Bernieri, Faulkner, Grahe, & Gada-Jain,2000). In contrast, withdrawal-oriented responses to exclusion (which represent a fairly consistentfinding within both correlational and experimental research) include: worse mood, increased anger/aggression and interpersonal deviance (Bierman & Wargo, 1995; Dodge, Coie, & Brakke, 1982;Hitlan, Kelly, Schepman, Schneider, & Zárate, 2006; Twenge, Baumeister, Tice, & Stucke, 2001;Williams, 2007), increased derogation of others (Bourgeois & Leary, 2001), and a desire to avoidfuture contact with perpetrators (Cheuk & Rosen, 1994; Leary, Koch, & Hechenbleikner, 2001;Pepitone & Wilpizeski, 1960). Such withdrawal-oriented responses, such as aggression, have alsobeen found to extend to individuals not directly involved in the experience (Twenge et al., 2001).Additionally, research on cyberostracism—which refers to instances of being rejected, ignored, orexcluded via some form of online social interaction—has found similar effects as those involvingostracism within face-to-face communications (Smith & Williams, 2004; Williams, Cheung, & Choi,2000; Williams et al., 2002).

Ostracism, language, and group interactions

Over the past few years, research has extended the conceptualization of exclusion and ostracism toinclude a language-based component. According to Hitlan et al. (2006), language-based exclusionrefers to any situation in which a target individual (or group) is ostracized by another individual (orgroup) via a language with which the target has extremely limited familiarity and understanding.

2 R. T. HITLAN ET AL.

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

Hitlan et al., randomly assigned participants to read one of three vignettes describing a series ofworkplace interactions between a protagonist and two other coworkers. Participants were asked toassume the perspective of the main character in the vignette. All participants read the sameintroduction indicating that the main character worked for a major auto manufacturer and washappy with his/her job, workgroup, and career opportunities. After the introduction participantsread that their other work-group members could speak both English and Spanish well. At this pointthe ostracism manipulation was administered. One third of the participants continued reading abouta series of interactions between themselves (the main character) and two work-group memberswherein the participant was actively included by their other group members over the course ofseveral different types of workplace interactions. A second third of participants read about how themain character was ostracized by his/her work-group members during the same series of work-related interactions even after repeated requests from the main character to stop (EnglishOstracism). The final third of participants were ostracized during work-group interactions in alanguage they did not know: Spanish (Spanish Ostracism).

It was predicted that ostracism, especially Spanish Ostracism (SO) would elicit group-basedcognitions (i.e., how similar the target felt toward his/her other group members) and facilitatethinking about others as either in-group or out-group members. Overall, ostracized participantsreported being less emotionally committed and feeling less obligated to their organization.Additionally, SO participants indicated significantly reduced workgroup commitment and reportedsignificantly greater symbolic threat (the degree to which participants felt that Mexican immigrationthreatens American culture and values) compared to EO participants. Moreover, SO resulted inhigher levels of expressed prejudice directed toward others not directly responsible for the behavior(i.e., general class of Mexican immigrants).

Related to the above research, Dotan-Eliaz, Sommer, and Rubin (2009) investigated linguis-tic ostracism. Linguistic ostracism was defined as, “any situation in which people converse in alanguage that others cannot understand” (p. 364). In this research, participants were eitherincluded or ostracized by two confederates who oscillated back and forth between speakingEnglish and Russian in the ostracism condition. Russian ostracized participants reportedhigher levels of rejection and anger, were less attracted to their coworkers, and viewed theirteam as less effective compared to linguistically included participants. Moreover, their percep-tion of their group’s performance was most negatively impacted when linguistically ostracizedparticipants also reported being more sensitive to such experiences (i.e., rejection sensitivity;Berenson et al., 2009; Downey & Feldman, 1996).

Consistent with these findings, other researchers have found that the effects of rejection andostracism are moderated by several individual difference variables including trait self-esteem anddepression, perceived control, social anxiety, and gender (for a review, see Williams, 2007). Takentogether, research investigating the role of individual differences in this area has been importantfor answering questions related to “when” and “for whom” rejection and/or ostracism is mostlikely to exert its effects; yet, it is also important to understand “why” such behavior impactstargets. That is, what happens during the process of being rejected or ostracized that influencessubsequent responses?

The current research was designed to extend our earlier work on language-based (i.e., linguistic)ostracism. The current research used a cyber-communication paradigm, and as such, the currentresearch sought to integrate aspects of both cyber- and linguistic ostracism into a single researchdesign. This area seems particularly important for better understanding the interplay betweentechnology and social behavior and extends the concept of language ostracism to situations involvingonline modes of communication (e.g., chatroom, discussion board).

When interacting with other people, the categorization of “others” into in-groups and out-groups is often ambiguous. One can use multiple dimensions to categorize others (Zárate &Smith, 1990), depending on the context. However, because language use represents an importantcomponent of one’s ethnic identity (Cargile & Giles, 1997, 1998; Giles et al., 1976, 1995), we

THE JOURNAL OF SOCIAL PSYCHOLOGY 3

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

argue that the language used to communicate among group members would serve to highlightdifferentiation cues and increase the likelihood of viewing and responding to others in terms ofan in-group and out-group classification scheme. We further expected that such a classificationscheme would serve to quickly highlight ethnicity and accentuate group-based distinctions relatedto ethnicity (Bargh, Chen, & Burrows, 1996; Haslam Oakes, Turner, & McCarty, 1995; Turner,Oakes, Haslam, & McCarty, 1994)

Related to this, Esses, Jackson, and Armstrong (1998) proposed an instrumental model of groupconflict which proposes that both resource competition and out-group salience are important factorsin determining attitudes toward immigrants and immigration. Salience refers to factors such as novelappearance and behavior. Research has found support for both competition and salience as impor-tant factors in determining attitudes toward immigrants (Zárate, Garcia, Garza, & Hitlan, 2004). Wehypothesize that linguistic ostracism will highlight group-based distinctions and facilitate thecategorization of others on the basis of ethnic differences. In doing so, we expected linguisticostracism to accentuate intergroup difference leading to lower levels of perceived similarity betweentargets and perpetrators.

Research on intergroup emotions also suggests that such categorizations will have additionaldownstream effects on emotional states and group-based prejudices. According to IntergroupEmotion Theory (IET; Mackie, Devos, & Smith, 2000; Miller, Smith, & Mackie, 2004), one’semotions are intrinsically linked to group membership. More specifically, “When social identityis salient, appraisals of situations or events related to social identity focus on social rather thanpersonal outcomes. When emotions occur on a group basis, such emotions are often experiencedon behalf of the in-group and the in-group and out-group become the targets of emotion”(Mackie et al., 2000, p. 603).

Consistent with this idea, Desteno, Dasgupta, Bartlett, and Cajdric (2004) found that currentemotional state was an important factor in activating and/or creating unconscious intergroup biases.In this research, participants were classified as being either “overestimators” or “underestimators”based on the results of a bogus personality test. Participants then completed a measure of automaticattitudes toward the “out-group” using the implicit association test (IAT) both prior to andimmediately after an emotion induction task. During the emotion induction task, participantswere instructed to write about a past event that had made them angry (as compared to a sad orneutral event). IAT results between pre- and post-emotion induction indicated that anger wascapable of producing an automatic bias (i.e., prejudice) against individuals previously classified asout-group members. We test the hypothesis that linguistic (i.e., Spanish) ostracism will elicitincreased levels of anger and prejudice toward those thought to be responsible for the ostracismand the larger out-group to which the perpetrators belong (Twenge et al., 2001). In doing so, thecurrent research begins to fill a gap in understanding the processes through which linguisticostracism may exert its effects.

Study overview and predictions

Within the current research, participants were randomly assigned to be either included or ostracizedby two other group members (research confederates) during an online group-based discussion. Inthe inclusion condition, participants were actively included during the entire task. In the ostracismconditions, participants were totally excluded from the group task after an initial introductoryperiod. Specifically, in the English Ostracism (EO) condition, participants were ostracized viaEnglish where the other group members directed all communication only to one another whileignoring any inclusionary attempts made by the participant. In the Spanish Ostracism (SO) condi-tions, participants were treated identically to the above, except that the other group membersswitched from communicating in English to communicating in Spanish after the introductoryperiod. After completing the online group-based discussion, participants completed several measuresassessing their perceptions of their group members and Mexican immigrants as a group.

4 R. T. HITLAN ET AL.

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

First, we expect that ostracism will reduce the perceived similarity between participants and theirgroup members, especially when exposed to SO due to its ability to accentuate group-baseddistinctions (Bargh, Chen, & Burrows, 1996; Haslam & Turner, 1992; Zárate & Smith, 1990). Wealso expect that, consistent with previous research on linguistic ostracism (Dotan-Eliaz, Sommer, &Rubin; 2009) and intergroup emotions (Cottrell & Neuberg, 2005; Mackie et al., 2000), a similar setof relations will emerge for anger, with ostracized participants reporting more anger than includedparticipants, especially when exposed to SO. Third, we expect that participants’ perceptions ofreduced similarity to their other group members would be an important factor in determininggroup-based emotional states (i.e., anger; Cottrell & Neuberg, 2005; Desteno et al., 2004) and predictthat perceived similarity will mediate the relation between ostracism and anger (i.e., ostracism >perceived similarity > anger). Fourth, in line with previous research indicating that linguisticostracism can elicit anger (Chow, Tiedens, & Govan, 2008; Dotan-Eliaz et al., 2009) and the abilityof anger to provoke prejudice (DeSteno et al., 2004; Parker-Tapias, Glaser, Keltner, Vasquez, &Wickens, 2007), we expect that the relation between SO and prejudice will be mediated by anger (i.e.,SO > anger > prejudice). Finally, we test the hypothesized full serial multiple mediation model fromSO > perceived similarity > anger > expressed prejudice.

Method

Participants

Participants included 84 students (56 women and 28 men) from a mid-sized Midwestern university,ranging in age from 18 to 27 years (M = 19.48, SD = 1.76). All participants received five dollars ascompensation for their participation. The majority of participants reported being Caucasian (68.3%)followed by African-American (22%), Latino/Mexican-American (3.7%), and Other (6.1%).

Procedure and materials

After providing informed consent, participants were told that the purpose of the study was toinvestigate the effectiveness of different modes of communication (e.g., chatroom, face-to-face,videoconferencing) on stimulating discussion. Participants were further instructed that they wouldbe communicating with other participants via an Internet chatroom. As part of the cover story,participants were informed that they would be participating in a group discussion with two otherstudents from different universities. In actuality, the other participants were research confederateswho interacted with the actual participant according to predetermined scripts. During the introduc-tion period (initial 3 minutes of the 15 minute group discussion period), the confederates introducedthemselves and always included the participant. If the participant did not volunteer introductoryinformation, one of the confederates asked the participant to introduce him/herself. At the end of the3-minute introductory period, the experimental manipulation was administered.

Participants were randomly assigned to one of four conditions: (1) inclusion, (2) English ostra-cism (EO), (3) Spanish ostracism (SO), or (4) Spanish ostracism comparison (SOC). For the firstthree conditions, the discussion topic of Mexican immigration to the United States was chosen tohighlight and strengthen the effectiveness of the group-based manipulation and to also act as anatural lead-in to discover that the confederates knew Spanish and to the dependent measures,which included attitudes towards immigration. In the last condition, SOC, participants were alsoostracized via Spanish, but the discussion topic was changed in order to account for potentialpriming effects associated with the immigration discussion topic.

Across all experimental conditions, the confederates logged into the chatroom using the samenames and university affiliations. In the inclusion condition, the participants were actively includedby the other group members throughout the entire discussion period. Confederates were polite, andreceptive to the participant’s suggestions, thoughts, and opinions. If participants were not engaged in

THE JOURNAL OF SOCIAL PSYCHOLOGY 5

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

the discussion, the confederates prompted the participant by asking about his/her thoughts andopinions. In the EO condition, the confederates directed the conversation to each other by specifyingthe name of the group member to whom a comment was directed. All attempts by the participant tocontribute were ignored by the other group members. The confederates did not reply to anyquestions, comments, or requests made by the participant. In both of the first two conditions(inclusion and EO), all conversation between group members took place in English.

For the SO and SOC conditions, the same procedure was followed as indicated above, except that,immediately after a 3-minute introductory period, one confederate indicated that his father was bornin Mexico and that his parents had taught him how to read and speak Spanish. After hearing this, thesecond confederate indicated that her parents were also born in Mexico and that she too could readand write Spanish. The confederates then began sending messages to each other in Spanish. Due tothe potential for the discussion topic to prime culturally sensitive responding, the discussion topic inthe SOC condition was on “The Freshman 15—Truth or Fiction.” The SOC condition followed thesame procedure, and the discussion followed a similar length and tone.

After completing the group discussion, participants completed a questionnaire that includedmeasures about their chatroom experiences, social attitudes, and prejudicial attitudes. In additionto those measures described below, we also collected information on the extent to which participantsidentified with being American (i.e., American identification) and social dominance orientation(Pratto, Sidanius, Stallworth, Malle, 1994; Sidanius & Pratto, 2001). However, because these were notdirectly related to the current set of predictions, these measures were not analyzed. After completingthe final questionnaire, participants were debriefed, probed for suspicion, and dismissed. During thedebriefing period, two participants indicated being suspicious about the true nature of the study. Asa result, these participants were removed from subsequent data analyses.

Measures

Manipulation checkTo determine the effectiveness of the exclusion manipulation, participants answered three questionsassessing how accepted/ostracized they felt during the group discussion (e.g., How accepted did youfeel by the other group members?, How much interaction did you have with the other group members?,and How cohesive was your group?; α = .88). All responses were obtained on 7-point scales rangingfrom 1 (Not at all) to 7 ((Extremely). Scale scores were created by averaging across scale items.Higher average scores indicate greater perceived acceptance by the other group members.

Perceived similarityThe degree to which participants perceived of themselves as being similar to their other groupmembers was assessed using four questions designed for the current research (e.g., How similar doyou feel you were to the other participants in the study?, How much did you like the other members ofyour group?; α = .89). All responses were obtained on 7-point scales ranging from 1 (Not at all) to 7(Extremely). Scale scores were created by averaging across scale items. Higher average scores indicategreater perceived similarity to the other group members.

Emotional state (anger)Although the current research was limited to predictions regarding the emotional state of anger, toavoid possible demand characteristics, several additional emotional state descriptors were includedin the assessment of emotion. These included adjectives tapping into 5 broad emotion categoriesderived from previous research via cluster analysis (Robinson & Clore, 2001; Storm & Storm, 1987).Participants were asked the extent to which each adjective was descriptive of their current state on a9-point scale with endpoints ranging from 1 (I feel none of this in response to the situation) to 9 (I feela lot of this in response to the situation). Positive and negative adjective descriptors were mixedacross all scale items. Responses to the emotion descriptors angry, irritated, and disgusted were

6 R. T. HITLAN ET AL.

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

combined to form a measure of anger (α = .76). Scale scores were created by averaging across scaleitems. Higher numbers indicate feeling more anger as the result of their current situation.

Expressed prejudicePrejudice was assessed with the expressed prejudice scale (Stephan, Ybarra, Martinez, Schwarzwald,& Tur-Kaspa, 1998). The prejudice items asked participants to rate their feelings about immigrantsacross 12 attitudinal items (e.g., admiration, dislike, acceptance, superiority, affection, disdain;α = .82). All responses were obtained on a 10-point scale ranging from 0 (None at all) to 9(Extreme). A scale score was created by reverse-scoring where necessary and averaging acrossitems. Higher scores indicate more prejudicial attitudes toward Mexican immigrants andimmigration.

Spanish language ability/fluencySpanish ability/fluency was measured by four items assessing different aspects of language fluency,including reading, writing, speaking, and listening (α = .91). Participants were asked to rate theirability to read, write, and speak Spanish. All responses were obtained on a 5-point scale ranging from1 (No ability) to 5 (Excellent ability). Higher average scores were indicative of better Spanishlanguage ability.

Results

Spanish ability/fluency

The mean score for Spanish fluency was 2.04 (SD = .85) indicating that, overall, participants reportedbeing “below average” in their ability to read, write, and understand the Spanish language. Due to itsimportant role in the current research, we also examined Spanish language ability for thoseparticipants who experienced SO. Two participants in the Spanish ostracism conditions self-reportedtheir average Spanish ability to be “3” (average) or higher. As a result, these two participants wereremoved from subsequent analyses. After removing these two participants, the new mean Spanishability/fluency was 1.99 (SD = .81). Given the primary SO manipulation focused on participants’ability to read Spanish we also examined participants’ specific ability to read Spanish. Overall, theparticipants reported their mean Spanish reading ability to be in the “below average” range(M = 2.11, SD = .94). In addition, it should also be noted that, students in rural Illinois generallyhave few opportunities to practice their Spanish, and we believe that the manipulation was effective.1

Spanish (linguistic) ostracism

In an attempt to simplify analysis and increase statistical power, while retaining the integrity of theostracism manipulation, we examined whether the two Spanish ostracism conditions could bemerged into a single SO condition. To this end, we examined the differences between the two SOconditions on the measures of acceptance, similarity, anger, and prejudice. No significant differ-ences emerged between the SO and SOC conditions. Additionally, results indicated effect sizes of,d = .19, d = .38, d = .11, and d = .02, for each of these measures respectively (Appendix Aprovides the means, standard deviations, Cohen’s d effect size estimates, and associated p values).Only the effect size for perceived similarity was different from zero based on effect size confidenceintervals. As a whole, we interpreted the overall pattern of findings as indicating the two SOgroups produced similar effects, and they were combined into a single SO condition for hypothesistesting.

THE JOURNAL OF SOCIAL PSYCHOLOGY 7

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

Manipulation check

Results of a One-Way Analysis of Variance (ANOVA) indicated a significant main effect ofexperimental condition on acceptance, F(2, 77) = 68.18, p < .001, η2 = .64. Follow-up tests(Tukey) indicated significant differences between each of the experimental conditions with SOparticipants reporting the lowest level of acceptance during the group task (M = 1.94, SD = .94),followed by EO participants (M = 2.57, SD = 1.22), and included participants (M = 5.00, SD = .97).Table 1 lists the means and standard deviations for each condition, effect size estimates, and p valuesfor all pairwise comparisons. These results are viewed as supporting the overall effectiveness of theexperimental manipulation.

Perceived similarity

Overall, perceived similarity was negatively related to both anger, r(79) = –.48, p < .001, andexpressed prejudice, r(79) = –.40, p < .001. The less similar participants felt they were to theirother group members, the more anger and expressed prejudice they reported. Additionally, con-sistent with predictions, results of a One-Way ANOVA indicated a significant main effect ofostracism on perceived similarity, F(2, 77) = 37.22, p < .001, η2 = .49. Follow-up tests indicatedsignificant differences between each of the experimental conditions with SO participants reportingthe lowest level of perceived similarity (M = 2.04, SD = 1.12), followed by EO participants (M = 2.92,SD = 1.43), and included participants (M = 5.01, SD = 1.05). Table 1 lists the means and standarddeviations for each condition, effect size estimates, and p values for all pairwise comparisons.

Emotional state (anger)

In addition to the relation between anger and perceived similarity noted above, anger was alsorelated to expressed prejudice, r(79) = .39, p < .001. Overall, participants who reported more angeralso expressed more prejudice. In addition, results of a One-Way ANOVA indicated a significantmain effect for ostracism on anger, F(2, 76) = 8.02, p = .001, η2 = .17. Follow-up tests indicated thatSO participants reported being significantly more angry (M = 4.46, SD = 2.37) than includedparticipants (M = 2.25, SD = 1.53). EO participants (M = 3.52, SD = 1.82) were not significantly

Table 1. Means, standard deviations, and effect size estimates for study variables as a function of inclusionary status.

Variable

Inclusion(1)

n = 20

English Ostracism(2)

n = 21

Spanish Ostracism(3)

n = 39

M SD M SD M SD dw [CI95%] p

Acceptance 5.00a .97 2.57b 1.22 1.94c .921/2 2.25 [1.924–2.584] <.0011/3 3.32 [3.088–3.558] <.0012/3 .62 [.363–.877] =.048Similarity 5.01a 1.05 2.92b 1.43 2.04c 1.121/2 1.70 [1.326–2.078] <.0011/3 2.75 [2.479–3.029] <.0012/3 .72 [.417–1.032] =.032Anger 2.25a 1.53 3.52ab 1.82 4.46b 2.371/2 .77 [.270–1.276] =.1121/3 1.06 [.523–1.591 <.0012/3 .44 [–.111–.982] =.204Prejudice 3.04a 1.15 3.35ab 1.28 3.96b 1.381/2 .26 [–.103–.625] =.7171/3 .72 [.388–1.044] =.0292/3 .46 [.126–.796] =.191

Note. Pairwise comparisons using Tukey Honestly Significant Difference (HSD) to correct for inflated Type 1 error rate. dw = Cohen’sd effect size estimate weighted by sample size. [CI95%] = 95% CI associated with the Cohen’s d point estimate.

8 R. T. HITLAN ET AL.

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

different from either included or SO participants on self-reported anger. However, further examina-tion of the 95% confidence intervals for the effect size point estimate comparing the inclusion to EOdid not include zero. This finding provides some additional support for the initial predictions.However, while the effect size for the difference between the EO and SO conditions was d = .44 andin the predicted direction, the CI surrounding the effect size point estimate included zero indicatingthis portion of the hypothesis failed to receive support (see Table 1).

Prejudice towards Mexican immigrants

As predicted, results of a One-Way ANOVA indicated a significant main effect of ostracism onprejudice toward Mexican immigrants, F(2, 77) = 3.83, p = .026, η2 = .09. Follow-up tests indicatedsignificant differences between the inclusion and SO condition, with SO participants expressingsignificantly higher levels of prejudice (M = 3.96, SD = 1.38) compared to included participants(M = 3.04, SD = 1.15). EO Participants (M = 3.35, SD = 1.28) were not significantly different fromeither the inclusion or SO conditions. However, again, the effect size for the difference between theEO and SO conditions was d = .46 and the CI did not include zero, providing some additionalsupport for the initial prediction that EO and SO would differ in expressed prejudice. Moreover, thiseffect size was almost twice as large as the difference between the inclusion and EO conditions(d = .25) (see Table 1).

Perceived similarity as a mediator of the SO > anger relation

Next, we tested the prediction that perceived similarity would mediate the effect of SO on angerusing the PROCESS macro for SPSS (Hayes, 2013). This simultaneous approach to mediation ismore powerful than other multi-step approaches for directly testing the significance of indirecteffects in mediation models. In addition, this approach uses a nonparametric inferential bootstrap-ping procedure (n = 10,000), which does not make any assumptions about the shape of the samplingdistribution underlying the indirect effect, to compute the bias-corrected 95% confidence intervalsassociated with the significance of the indirect effects (i.e., mediational effect). For all mediationanalyses, the focus was on differences between the inclusion and SO conditions.

Results indicated that the total effect of SO on anger (i.e., the effect of SO on anger notconsidering perceived similarity) was significant, b = 1.56, SE = .46, t = 3.39, p = .001. Consistentwith predictions, after controlling for the effect of the proposed mediator—perceived similarity—thedirect effect of SO on anger was no longer significant, b = .58, SE = .53, t = 1.09, p = .280.Importantly, the indirect effect (i.e., mediated effect) of SO on anger through similarity wassignificant, b = .98, BootSE = .41, BootCI95 = [.2675–1.8504], κ2 = .20, BootSE = .07,BootCI95 = [.0613, .3453]. These findings provide evidence that perceived similarity mediates theeffect of SO on anger.

Anger as a mediator of the SO > prejudice relation

Next, we tested the prediction that anger would mediate the effect of SO on prejudice using theprocedure identical to that notes above. Results indicated that the total effect of SO on prejudice (i.e.,the effect of ostracism on prejudice not considering anger) was significant, b = .81, SE = .29, t = 2.82,p = .006. However, after controlling for the effect of the proposed mediator—anger—the direct effectof SO on prejudice was no longer significant, b = .50, SE = .29, t = 1.70, p = .092). Of import, theindirect effect (i.e., mediated effect) of SO on prejudice through anger was significant, b = .31,BootSE = .17, BootCI95 = [.0650, .7579]; κ2 = .11, BootSE = .06, BootCI95 = [.0241, .2735]. Thesefindings illustrate that, consistent with predictions, anger mediates the effects of SO on prejudice.

THE JOURNAL OF SOCIAL PSYCHOLOGY 9

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

Serial multiple mediator model

Finally, we tested the full serial multiple mediation model from linguistic ostracism > perceivedsimilarity > anger > prejudice. This model allows for the simultaneous testing of the indirect effectthrough both mediators and through each mediator by itself (i.e., controlling for the other mediatorin the model). Full model coefficients are presented in Figure 1. Results indicated a significant totaleffect for SO on prejudice, b = .44, SE = .18, t = 2.49, p = .015. The total indirect effect (i.e., totalmediation effect including both mediators) was also significant, b = .42, BootSE = .21,BootCI95 = [.0058, .8505]. To more fully assess the total indirect effect, we examined the contributionfor each mediator separately and together in serial fashion (i.e., specific indirect effects). Both of thespecific indirect paths through either perceived similarity alone, b = .21, BootSE = .25,BootCI95 = [–.1995–.7803] and anger alone, b = .03, BootSE = .12, BootCI95 = [–.1377, .3577] werenot significant. These findings suggest that neither perceived similarity nor anger alone wereindependent mediators of the effect of linguistic ostracism on prejudice. However, the serialmediation indirect effect path was significant, b = .17, BootSE = .13, BootCI95 = [.0022, .5617],providing support for a multi-step serial mediation effect from linguistic ostracism > perceivedsimilarity > anger > prejudice.2

Discussion

In the current research, participants interacted with other group members via a common mode ofonline communication, a chatroom. Consistent with theories on intergroup relations, we expectedthat ostracized participants would appraise themselves as being less similar to their group members,especially when exposed to SO. Consistent with predictions, ostracized participants reported feelingless similar compared to their included counterparts. In addition, SO participants reported feelingsignificantly less similar compared to EO participants.

We also expected that ostracized participants would report being angrier than included partici-pants and that this effect would be strongest for SO participants. This hypothesis received modestsupport. SO participants reported being angrier compared to included participants. In addition,although the predicted difference in anger between the inclusion and EO conditions and between theEO and SO conditions were not statistically significant, the associated effect sizes for each wasd = .77 and d = .44 (in the predicted directions), respectively. The lack of statistical significance mayhave been an artifact of the study being underpowered. For example, post-hoc power analysis usingGpower 3.1.9.2 indicated a statistical power level of .49 (assuming a medium effect size) and .89(assuming a large effect size). Because it took two experimenters and two confederates to run 1participant, a larger sample was prohibitively expensive. Moreover, the overall pattern of effect sizesrelated to similarity, anger, and prejudice suggest the psychological effects of SO are more pro-nounced compared to when participants were ostracized in a known language. These findings are

SpanishOstracism

Perceived Similarity Anger

Expressed Prejudice

.44* (.03)

–1.49***

–.59*

.19*

.15–.14

Figure 1. Serial mediation model of the effect of linguistic ostracism on prejudice. Note. *p < .05, **p < .01, ***p < .001. Numericvalues represent regression coefficients. For the relationship between Spanish Ostracism (SO) and expressed prejudice, thecoefficient associated with the total effect of SO on prejudice is not in parentheses and the coefficient for direct effect of SOon prejudice (after including perceived similarity and anger in the model) is provided within parentheses.

10 R. T. HITLAN ET AL.

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

not only of theoretical interest but practically relevant to any society with immigrants that speak alanguage that is unfamiliar to citizens of the host country.

It was further predicted that SO, by virtue of its ability to highlight group-based distinc-tions, would lead to increased levels of prejudice toward Mexican immigrants. This predictionreceived partial support. SO participants expressed significantly higher levels of prejudicecompared to included participants. As was the case with anger, a non-significant trendemerged with SO participants reporting higher levels of prejudice compared to EO participants(d = .46). We expect that a similar power issue may be involved. Given this, we relied on bothsignificance testing and effect size estimates (with associated 95% CI’s) in determining themost likely pattern of relations.

Moreover, based on theories and research on intergroup conflict and emotions (Cottrell &Neuberg, 2005; Esses et al., 1998; Mackie & Smith, 2003; Mackie et al., 2000; Miller et al., 2004),it was predicted that: (1) the effect of SO on anger would be mediated by perceived similarity;(2) the effect of SO on prejudice would be mediated by anger; and (3) perceived similarity andanger would serially mediate the relation between SO and prejudice. Each of these hypotheseswas supported. Both similarity and anger were found to be significant mediators of theirrespective effects. Additionally, when both similarity and anger were included in the final fullserial mediation model (which examines both the independent and combined mediation effects)neither similarity nor anger were found to act independently; rather, these variables actedconjointly in providing support for a multi-step serial mediation effect. This may be a functionof the correlation/shared variance between these two variables, r(79) = –.48, p < .001, that isaccounted for within the full serial mediation model.

The current research represents an important step in better understanding how language usecan shape peoples’ attitudes. Arguably, language use has become an important symbolic andpolitical issue in the United States. Given the changing demographics and increased Latinopresence in the United States, language use will continue to be an important dimension onwhich people are categorized, judged, and evaluated. Consistent with this, the current researchalso supports some basic postulates underlying intergroup emotions theory. For example,manipulations intended to highlight group boundaries were associated with higher levels ofanger and prejudice. In the broader context of intergroup relations, previous research suggeststhat being ostracized by similar others (e.g., in-group members) elicits physiological and biolo-gical responses consistent with a challenge response as indexed by higher cardiac output andlower total peripheral resistance (Blascovich, Berry-Mendes, & Seery, 2003) and increasedcirculating testosterone levels (DeSoto, Hitlan, Deol, & McAdams, 2010); whereas, being ostra-cized by dissimilar others has been linked to decreased circulating levels of free testosterone(DeSoto et al., 2010) and increased circulating levels of free cortisol (DeSoto & Hitlan, 2015).

As a whole, the research on social ostracism and rejection, along with the intergroupdistinctions noted above, provide evidence that being ostracized or otherwise rejected isaversive to targets. However, the motivational tendencies underlying how one copes withsuch experiences seem contingent on whether the behavior (or lack thereof) is being perpe-trated by others perceived to be similar (e.g., in-group members) or dissimilar (e.g., out-groupmembers) to oneself. It appears social threats, including threats to one’s social status, elicitmotivational tendencies consistent with a challenge or threat response depending on perpe-trator characteristics (Blascovich et al., 2003; DeSoto & Hitlan, 2015; DeSoto et al., 2010)

Limitations and directions for future research

Due to the time-consuming nature of the current research, some practical limitations were evident.First, while the current research had ample power for detecting large effects, the power to detectmoderate and small effect size estimates was substantially reduced. Larger (and more diverse)samples would help to establish the statistical significance of group differences via traditional null

THE JOURNAL OF SOCIAL PSYCHOLOGY 11

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

hypothesis significance testing (Cohen, 1992). Nevertheless, the current research represents one ofthe first empirical studies examining linguistic ostracism and how linguistic ostracism compares tomore traditional non-linguistic ostracism situations. We also believe this to be some of the firstresearch to test a serial multiple mediator model through which SO exerts its effects on downstreamresponses/outcomes.

In addition, this initial study was designed to use only Spanish as the linguistic ostracismlanguage. Future research may also want to consider the status of the linguistic ostracism languageand examine how language status influences this entire process. Presumably, Mexicans and immi-grants are often thought to represent lower status groups by the prototype “White” American.Would exclusion from a high status group produce the same or similar effects? Thus, Swedish, forexample, might be considered more “exotic” and produce more curiosity than anger. Future researchcould also seek to investigate how and to what extent the use of different languages is related todifferent types of threat. It is plausible that individuals learn to cope in a multilingual world. So, towhat extent are the current findings reproducible and are the current findings unique to youngerAmerican adults who are monolingual English speakers? The sample was picked because it is awayfrom larger city centers that might have a more diverse population. One might predict that theseeffects occur in more homogenous small towns, though that question remains open to empirical test.It would also be interesting to investigate if the obtained or similar effects emerge when examiningthe linguistic ostracism of minorities.

Future research in this area may also benefit from manipulating the discussion topic. Itremains unclear if other “more neutral” discussions or work-related topics would engendersimilar group-based responses. Moreover, while the current research represents an importantstep in better understanding the path through which linguistic ostracism (Spanish) impactsexpressed prejudice, a more robust test may involve measuring prejudice both before and aftersuch interactions. This way initial prejudice levels could be used to investigate change or as apotential moderator of the relationships between linguistic ostracism (e.g., Spanish ostracism) andsimilarity, anger, and prejudice. In addition, the current research was not designed to examinethe link between prejudicial attitudes and actual manifest behavior. Thus, an important extensionof this research would involve understanding if increased prejudicial attitudes lead to somedownstream behavioral consequence (e.g., discrimination) to see how these ostracized (andnow angrier) participants actually behave.

Conclusion

In conclusion, because individuals are increasingly likely to encounter people from other cultures,there is a very real possibility of being ostracized via language, whether intentional or not. Although,such differences in perceived intent may themselves lead to different reactions based on theperpetrators perceived motivation (Williams, 2007). We found the experience of SO makes peopleangry and sometimes causes them to reveal prejudice, and these outcomes can be damaging tointergroup relations in a variety of different situations and contexts. (Hitlan et al., 2006). But, why doMexican immigrants provoke such anger and threat in monolingual English-speaking youngAmericans? Our results suggest that people can be angered by associated language differences.Being ostracized by others speaking an unknown/unfamiliar language, seems to engender strongercues consistent with a potential threat, signaling a stronger phenomenological experience, comparedto when people are ostracized via a known/familiar language.

Notes

1. Although, at the group level, participants mean Spanish reading ability was in the “below average” range, At theindividual level, there were 4 participants who reported their ability to read Spanish as “above average” and 1participant who reported their ability as “Excellent”. Even so, research suggests self-reports to be upwardly

12 R. T. HITLAN ET AL.

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

biased when assessing one’s ability (Sheppard, Malone, & Sweeny, 2008) and the inclusion of participants whoare better able to read Spanish should only work against our initial predictions.

2. We also tested an alternative serial mediation model by transposing the two serial mediators resulting in thefollowing serial mediation model: SO > Anger >Perceived Similarity > Expressed Prejudice. Results againsupported the overall mediation effect, with a statistically significant total indirect effect, b = .46, BootSE = .17,BootCI95 = [.1315, .8115]. Additionally, while the specific indirect path through anger alone was statisticallysignificant, b = .18, BootSE = .11, BootCI95 = [.0008, .4507], the serial mediation indirect effect path was notstatistically significant, b = .04, BootSE = .03, BootCI95 = [–.0011, .1345].

Notes on contributors

Robert Thomas Hitlan is a Professor of Psychology at the University of Northern Iowa and Director of thePsychoneuroendocrinology Lab. His research interests include understanding the psychological, physical, and biolo-gical correlates and consequences of social and workplace stressors including ostracism and harassment. Michael A.Zárate is a Professor of Psychology at the University of Texas at El Paso. One of Dr. Zárate’s main areas of researchinvolves developing a ‘‘cultural inertia’’ model of prejudice reduction using social identity concepts to test how ahypothesized fear of change can influence prejudice, particularly concerning immigrants. Kristine M. Kelly earned herPhD in experimental psychology from the University of Tennessee and is currently Professor of Psychology at WesternIllinois University. She is a former Honors College Associate Director and is a recipient of the university’s OutstandingResearch with Students award. M. Catherine DeSoto is a Professor of Psychology at the University of Northern Iowa.Her principal area of research involves investigating the influence of hormones on both normal behavior and theexpression of psychopathology, particularly borderline personality and neuroticism.

References

Associated Press (2000, August 16). Women sues Albertsons, charges racial discrimination. Amarillo Globe News.Retrieved from http://amarillo.com/stories/081600/tex_albertsons.shtml

Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct effects of trait construct andstereotype activation on action. Journal of Personality and Social Psychology, 71, 230–244. doi:10.1037/0022-3514.71.2.230

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamentalhuman motivation. Psychological Bulletin, 117, 497–529. doi:10.1037/0033-2909.117.3.497

Berenson, K. R., Gyurak, A., Ayduk, Ö., Downey, G., Garner, M. J., Mogg, K. . . . Pine, D. S. (2009). Rejectionsensitivity and disruption of attention by social threat cues. Journal of Research in Personality, 43, 1064–1072.doi:10.1016/j.jrp.2009.07.007

Bernstein, M. J., & Claypool, H. M. (2012a). Not all social exclusions are created equal: Emotional distress followingsocial exclusion is moderated by exclusion paradigm. Social Influence, 7, 113–130. doi:10.1080/15534510.2012.664326

Bernstein, M. J., & Claypool, H. M. (2012b). Social exclusion and pain sensitivity: Why exclusion sometimes hurts andsometimes numbs. Personality and Social Psychology Bulletin, 38, 185–196. doi:10.1177/0146167211422449

Bierman, K. L., & Wargo, J. B. (1995). Predicting the longitudinal course associated with aggressive-rejected, aggressive(nonrejected), and rejected (nonaggressive) status. Development and Psychopathology, 7, 669–682. doi:10.1017/S0954579400006775

Blascovich, J., Berry-Mendes, W., & Seery, M. D. (2003). Intergroup encounters and threat: A multimethod approach.In D. M. Mackie & E. R. Smith (Eds.), From prejudice to intergroup emotions: Differentiated reactions to socialgroups (pp. 89–109). New York, NY: Psychology Press.

Borgeois, K. S., & Leary, M. R. (2001). Coping with rejection: Derogating those who choose us last. Motivation andEmotion, 24, 101–111. doi:10.1023/A:1010661825137

Cargile, A. C., & Giles, H. (1997). Understanding language attitudes: Exploring listener affect and identity. Language &Communication, 17, 195–217. doi:10.1016/S0271-5309(97)00016-5

Cargile, A. C., & Giles, H. (1998). Language attitudes toward varieties of english: An american-japanese context.Journal of Applied Communication Research, 26, 338–356. doi:10.1080/00909889809365511

Cheuk, W. H., & Rosen, S. (1994). Validating a “spurning scale” for teachers. Current Psychology, 13, 241–247.doi:10.1007/BF02686851

Chow, R. M., Tiedens, L. Z., & Govan, C. (2008). Excluded emotions: The role of anger in antisocial responses toostracism. Journal of Experimental Social Psychology, 44, 896–903. doi:10.1016/j.jesp.2007.09.004

Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159. doi:10.1037/0033-2909.112.1.155Contreras, R. (2013, June 6). Whole foods workers say they were suspended for speaking Spanish. Associated Press. Retrieved

from http://nbclatino.com/2013/06/06/whole-foods-workers-say-they-were-suspended-for-speaking-spanish/

THE JOURNAL OF SOCIAL PSYCHOLOGY 13

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

Cottrell, C. A., & Neuberg, S. L. (2005). Different emotional reactions to different groups: A sociofunctional threat-based approach to “Prejudice”. Journal of Personality and Social Psychology, 88, 770–789. doi:10.1037/0022-3514.88.5.770

DeSoto, M. C., & Hitlan, R. T. (2015). Can dual hormones ever duel? Neuroendocrine interactions during socialexclusion. North American Journal of Psychology 17, 555–574.

DeSoto, M. C., Hitlan, R. T., Deol, R.-S.-S., & McAdams, D. (2010). Testosterone fluctuations in young men: Thedifference between interacting with like and not-like others. Evolutionary Psychology, 8, 173–178. doi:10.1177/147470491000800203

Desteno, D., Dasgupta, N., Bartlett, M. Y., & Cajdric, A. (2004). Prejudice from thin air. The effect of emotion onautomatic intergroup attitudes. Psychological Science, 15, 319–324. doi:10.1111/psci.2004.15.issue-5

Dodge, K. A., Coie, J. D., & Brakke, N. P. (1982). Behavior patterns of socially rejected and neglected preadolescents:The roles of social approach and aggression. Journal of Abnormal Child Psychology, 10, 389–409. doi:10.1007/BF00912329

Dotan-Eliaz, O., Sommer, K., & Rubin, Y. (2009). Multilingual groups: Effects of linguistic ostracism on felt rejectionand anger, coworker attraction, perceived team potency, and creative performance. Basic and Applied SocialPsychology, 31, 363–375. doi:10.1080/01973530903317177

Downey, G., & Feldman, S. I. (1996). Implications of rejection sensitivity for intimate relationships. Journal ofPersonality and Social Psychology, 70, 1327–1343. doi:10.1037/0022-3514.70.6.1327

Esses, V. M., Jackson, L. M., & Armstrong, T. L. (1998). Intergroup competition and attitudes toward immigrants andimmigration: An instrumental model of group conflict. Journal of Social Issues, 54, 699–724. doi:10.1111/j.1540-4560.1998.tb01244.x

Fuchs, E. (2008, November 6). Salvation army, EEOC Settle English-Only Suit. Law 360, Retrieved from http://www.law360.com/articles/75789/salvation-army-eeoc-settle-english-only-suit

Garcia v. Gloor, 618 F.2d 264 (5th Cir. 1980, cert denied, 449 U.S. 1113 (1981).Giles, H., Taylor, D. M., Lambert, W. E., & Albert, G. (1976). Dimensions of ethnic identity: An example from

northern Maine. The Journal of Social Psychology, 100, 11–19. doi:10.1080/00224545.1976.9711902Giles, H., Williams, A., Mackie, D. M., & Rosselli, F. (1995). Reactions to Anglo- and Hispanic-American-accented

speakers: Affect, identity, persuasion, and the English-only controversy. Language & Communication, 15, 107–120.doi:10.1016/0271-5309(94)00019-9

Haslam, A., & Turner, J. (1992). Context dependent variation in social stereotyping2: The relationship between frameof reference, self-categorization, and accentuation. European Journal of Social Psychology, 22, 251–277. doi:10.1002/(ISSN)1099-0992

Haslam, S. A., Oakes, P. J., Turner, J. C., & McCarty, C. (1995). Social categorization and group homogeneity: Changesin the perceived applicability of stereotype context as a function of comparative context and trait favourableness.British Journal of Social Psychology, 34, 139–160. doi:10.1111/j.2044-8309.1995.tb01054.x

Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis. New York, NY: Guilford.Hitlan, R. T., Kelly, K. M., Schepman, S., Schneider, K. T., & Zárate, M. A. (2006). Language exclusion and the

consequences of perceived ostracism in the workplace. Group Dynamics: Theory, Research, and Practice, 10, 56–70.doi:10.1037/1089-2699.10.1.56

Leary, M. R., Koch, E. J., & Hechenbleikner, N. R. (2001). Emotional responses to interpersonal rejection. In M. R.Leary (Ed.), Interpersonal rejection. New York, NY: Oxford University Press.

Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emotions: Explaining offensive action tendencies in anintergroup context. Journal of Personality and Social Psychology, 79, 602–616. doi:10.1037/0022-3514.79.4.602

Mackie, D. M., & Smith, E. R. (Eds.). (2003). From prejudice to intergroup emotions: Differentiated reactions to socialgroups. Philadelphia, PA: Psychology Press.

Malkin, M. (2007). The forgotten “A” word: Assimilation. Retrieved from http://townhall.com/columnists/MichelleMalkin/2007/07/04/the_forgotten_a_word_assimilation.

Miller, D. A., Smith, E. R., & Mackie, D. M. (2004). Effects of intergroup contact and political predispositions onprejudice: Role of intergroup emotions. Group Processes and Intergroup Relations, 7, 221–237. doi:10.1177/1368430204046109

Molden, D. C., Lucas, G. M., Gardner, W. L., Dean, K., & Knowles, M. L. (2009). Motivations for prevention orpromotion following social exclusion: Being rejected versus being ignored. Journal of Personality and SocialPsychology, 96, 415–431. doi:10.1037/a0012958

Parker-Tapias, M., Glaser, J., Keltner, D., Vasquez, K., & Wickens, T. (2007). Emotion and prejudice: Specific emotionstoward outgroups. Group Processes & Intergroup Relations, 10, 27–39. doi:10.1177/1368430207071338

Pepitone, A., & Wilpizeski, C. (1960). Some consequences of experimental rejection. The Journal of Abnormal andSocial Psychology, 60, 359–364. doi:10.1037/h0042405

Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A personality variablepredicting social and political attitudes. Journal of Personality and Social Psychology, 67, 741–763. doi:10.1037/0022-3514.67.4.741

14 R. T. HITLAN ET AL.

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16

Robinson, M. D., & Clore, G. L. (2001). Simulation, scenarios, and emotional appraisal: Testing the convergence of realand imagined reactions to emotional stimuli. Personality and Social Psychology Bulletin, 27, 1520–1532. doi:10.1177/01461672012711012

Ryan, C. (2013). U.S. Bureau of the Census. Language use in the United States: 2011. Retrieved from http://www.census.gov/prod/2013pubs/acs-22.pdf

Shepperd, J., Malone, W., & Sweeny, K. (2008). Exploring causes of the self-serving bias. Social and PersonalityPsychology Compass, 2, 895–908. doi:10.1111/j.1751-9004.2008.00078.x

Sidanius, J., & Pratto, F. (2001). Social dominance: An intergroup theory of social hierarchy and oppression. Cambridge,UK: Cambridge University Press.

Smart-Richman, L., & Leary, M. R. (2009). Reactions to discrimination, stigmatization, ostracism, and other forms ofinterpersonal rejection: A multimotive model. Psychological Review, 116, 365–383. doi:10.1037/a0015250

Smith, A., & Williams, K. D. (2004). R U There? Effects of ostracism by cell phone messages. Group Dynamics: Theory,Research, and Practice, 8, 291–301. doi:10.1037/1089-2699.8.4.291

Stephan, W. G., Ybarra, O., Martinez, C. M., Schwarzwald, J., & Tur-Kaspa, M. (1998). Threat towards immigrants toSpain and Israel: An integrated threat theory analysis. Journal of Cross-Cultural Psychology, 29, 559–576.doi:10.1177/0022022198294004

Storm, C., & Storm, T. (1987). A taxonomic study of the vocabulary of emotions. Journal of Personality and SocialPsychology, 53, 805–816. doi:10.1037/0022-3514.53.4.805

Turner, J. C., Oakes, P. J., Haslam, S. A., & McCarty, C. (1994). Self and collective: Cognition in social context.Personality and Social Psychology Bulletin, 20, 454–463. doi:10.1177/0146167294205002

Twenge, J. M., Baumeister, R. F., Tice, D. M., & Stucke, T. S. (2001). If you can’t join them, beat them: The effects ofsocial exclusion on antisocial behavior. Journal of Personality and Social Psychology, 81, 1058–1069. doi:10.1037/0022-3514.81.6.1058

US English Inc. (2015). Official English. Retrieved from http://us-english.org/view/24.Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425–452. doi:10.1146/annurev.

psych.58.110405.085641Williams, K. D., Bernieri, F., Faulkner, S., Grahe, J., & Gada-Jain, N. (2000). The scarlett letter study: Five days of social

ostracism. Journal of Personal and Interpersonal Loss, 5, 19–63. doi:10.1080/10811440008407846Williams, K. D., Cheung, C. K. T., & Choi, W. (2000). Cyberostracism: Effects of being ignored over the Internet.

Journal of Personality and Social Psychology, 79, 748–762. doi:10.1037/0022-3514.79.5.748Williams, K. D., Govan, C. L., Croker, V., Tynan, D., Cruickshank, M., & Lam, A. (2002). Investigations into

differences between social- and cyberostracism. Group Dynamics: Theory, Research, and Practice, 6, 65–77.doi:10.1037/1089-2699.6.1.65

Williams, K. D., & Sommer, K. (1997). Social ostracism by coworkers: Does rejection lead to loafing or compensation?Personality and Social Psychology Bulletin, 23, 693–706. doi:10.1177/0146167297237003

Zárate, M., Shaw, M., Marquez, J., & Biagas, D. (2012). Cultural inertia: The effects of cultural change on intergrouprelations and the self-concept. Journal of Experimental Social Psychology, 48, 634–645. doi:10.1016/j.jesp.2011.12.014

Zárate, M. A., Garcia, B., Garza, A. A., & Hitlan, R. T. (2004). Cultural threat and perceived realistic groupconflict as dual predictors of prejudice. Journal of Experimental Social Psychology, 40, 99–105. doi:10.1016/S0022-1031(03)00067-2

Zárate, M. A., & Smith, E. R. (1990). Person categorization and stereotyping. Social Cognition, 8, 161–185. doi:10.1521/soco.1990.8.2.161

Appendix. Means, Standard Deviations, p values, and effect size estimates for the two Spanish ostracism conditions.

Spanish Ostracism: SO (Immigration)(n = 20)

Spanish Ostracism: SOC (Freshman 15)(n = 21)

Variable M SD M SD t p dw [CI95%]

Acceptance 2.02 .91 1.88 .60 .56 .58 .19 [–.042—.614]Similarity 2.29 1.44 1.84 .96 1.17 .25 .38 [.015–.742]Anger 4.53 2.23 4.28 2.45 .34 .74 .11 [–.591–.809]Prejudice 3.90 1.44 3.93 1.38 –.06 .95 .02 [–.443–.399]

Note. dw = Cohen’s d effect size estimate weighted by sample size. [CI95%] = 95% CI associated with the Cohen’s d point estimate.

THE JOURNAL OF SOCIAL PSYCHOLOGY 15

Dow

nloa

ded

by [

Rod

Lib

rary

, Uni

vers

ity o

f N

orth

ern

Iow

a] a

t 10:

56 1

3 Ja

nuar

y 20

16