Psychology of Religion and Spirituality

Perceptions of Discrimination Among Atheists: Consequences for Atheist Identification, Psychological and Physical Well-Being Michael J. Doane and Marta Elliott Online First Publication, November 3, 2014. http://dx.doi.org/10.1037/rel0000015

CITATION Doane, M. J., & Elliott, M. (2014, November 3). Perceptions of Discrimination Among Atheists: Consequences for Atheist Identification, Psychological and Physical Well-Being. Psychology of Religion and Spirituality. Advance online publication. http://dx.doi.org/10.1037/rel0000015 Psychology of Religion and Spirituality © 2014 American Psychological Association 2014, Vol. 6, No. 4, 000 1941-1022/14/$12.00 http://dx.doi.org/10.1037/rel0000015

Perceptions of Discrimination Among Atheists: Consequences for Atheist Identification, Psychological and Physical Well-Being

Michael J. Doane and Marta Elliott University of Nevada, Reno

Atheists are a marginalized group of people in the United States. Although studies have found that atheists perceive prejudice and experience discrimination, little is known about the consequences of such social rejection. To address this gap in research, we examined the associations among discrimination, identity, and well-being using original survey data from self-identified atheists (n ϭ 960). We investi- gated their associations in the context of the Rejection–Identification Model, which posits that group identification reduces the negative effect of discrimination on well-being. Consistent with extant research on other marginalized groups, discrimination was negatively associated with well-being while positively associated with atheist identification. Further, atheist identification was positively associated with well-being. In support of the rejection–identification process, we found evidence that atheists may strengthen their group identification in the face of discrimination. Strengthened identification as an atheist may be a strategy that protects atheists from the harmful effects of social rejection.

Keywords: atheism, discrimination, identity, rejection–identification model, well-being

Atheists are one of the most stigmatized groups of people in the more strongly with their stigmatized status in the face of discrim- United States. Many people hold negative stereotypes of atheists, ination, a coping strategy that has the potential to reduce the costs believing them to be judgmental, cynical, and hedonistic (Harper, of discrimination for well-being (Branscombe, Schmitt, & Harvey, 2007). People also harbor negative feelings toward atheists, in- 1999). cluding distrust (Gervais, Shariff, & Norenzayan, 2011), disgust (Ritter & Preston, 2011), and fear (Franks & Scherr, in press). Who Are Atheists? Such negative feelings may translate into discriminatory behav- iors, such as slander and ostracism (Hammer, Cragun, Hwang, & Atheists are people who lack any belief in god(s) or explicitly Smith, 2012). Although recent research has documented prejudice believe in the nonexistence of god(s) (Cliteur, 2009). Many athe- (Gervais, 2013; Johnson, Rowatt, & LaBouff, 2012) and discrim- ists find religious beliefs and practices to be meaningless in their ination (Gervais et al., 2011; Hammer et al., 2012) targeting lives and inconsistent with their secular worldviews (Zuckerman, atheists, the consequences of such social rejection for atheists 2009). A hallmark of contemporary or “new” atheism is that themselves remain unclear. For example, is discrimination associ- atheists are hostile toward organized religion (e.g., Streib & Klein, ated with lower well-being among atheists? If so, how might 2013), yet not all atheists report antireligious sentiments (Cimino atheists protect themselves against the harmful effects of discrim- & Smith, 2011). In recent years, atheists have become more active ination? The purpose of this study was to address these two and visible in the United States (Guenther, Mulligan, & Papp, research questions. Drawing on research among other maligned 2013). For example, several prominent atheists have authored social groups, we expected that atheists’ perceptions of discrimi- books challenging religion and promoting atheism (e.g., Dawkins, nation would be associated with poor well-being, both psycholog- 2006; Hitchens, 2007). There has also been an increase in both ical and physical. We also anticipated that atheists would identify informal and formal atheist organizations, some of which are politically active at the local and national levels (Guenther et al., This document is copyrighted by the American Psychological Association or one of its allied publishers. 2013; Smith & Cimino, 2012). This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Researchers have generally defined atheists in one of two ways: (a) theological atheists, people who do not believe in god(s), and Michael J. Doane, Interdisciplinary PhD Program in Social Psychology, (b) self-identifying atheists, people who identify as atheist from University of Nevada, Reno; Marta Elliott, Department of Sociology, among other nonreligious categories. Theological and self- University of Nevada, Reno. identifying atheism frequently co-occur, but not always. For ex- The authors acknowledge their appreciation for the assistance of Atheist ample, some theological atheists choose to identify with other Alliance International, especially for the support from Carlos A. Diaz and nonreligious categories (e.g., humanist or secularist; Sherkat, Tanya Smith, in addition to the other AAI Directors, members, and 2008). In addition, some self-identifying atheists report varied supporters. The authors also thank Joseph H. Hammer for his suggestions on an earlier version of the survey used in this study. beliefs in god(s) (Cragun, Kosmin, Keysar, Hammer, & Nielsen, Correspondence concerning this article should be addressed to Michael 2012). This distinction between endorsement of in-group beliefs J. Doane, Interdisciplinary PhD Program in Social Psychology, University and self-identification with the in-group applies to many other of Nevada, Reno, Mail Stop 300, Reno, NV 89557. E-mail: mdoane@ social groups, such as the difference between religious beliefs and unr.edu identity among Jews (Dubow, Pargament, Boxer, & Tarakeshwar,

1 2 DOANE AND ELLIOTT

2000). We defined atheists for this study as people who satisfy friends, and coworkers) and involving such acts as ostracism (e.g., both theological and self-identifying atheism, meaning that they being avoided or isolated), coercion (e.g., pressured to engage in explicitly state that they do not believe in god(s) and identify religious activities), and slander (e.g., being told that one is sinful themselves using the label “atheist.” and immoral; Hammer et al., 2012). Anecdotal evidence also Large-scale surveys show that atheists tend to be male (Keysar, suggests that social rejection is a common experience among 2007) and relatively young in age (Pew Research Center, 2008). atheists (Hunsberger & Altemeyer, 2006). Further, even if atheists Nonreligious people in general tend to be educated (Kosmin & have never personally experienced discrimination, they likely rec- Keysar, 2006), politically liberal (Hout & Fischer, 2002), and less ognize their marginalized position in the United States (Cimino & likely to live in Southern states or rural areas (Sherkat, 2008). Smith, 2011). Although the number of people identifying as atheist—and non- Although discrimination is a particularly harmful social stressor religious, more broadly—is on the rise (Pew Research Center, for members of many marginalized groups (Weber, Pargament, 2012), these people represent a relatively small portion of the Kunik, Lomax, & Stanley, 2012), it is unclear whether experiences United States population. Based on nationally representative sam- of rejection undermine well-being for atheists. The broader liter- ples, estimates place self-identified atheists between 0.7% (Kos- ature indicates that belonging to a marginalized group comes with min & Keysar, 2009) and 2.4% (Pew Research Center, 2012) of psychological and even physical costs (Clark, Anderson, Clark, & the total United States population. These estimates are similar to Williams, 1999; Pascoe & Smart Richman, 2009). For example, those for many minority religious groups, including Mormons perceptions of discrimination are associated with increased symp- (1.7%) and Jews (1.7%; Pew Research Center, 2008). Figures for toms of anxiety and depression among African Americans (Banks, atheists may actually be conservative, as some people who pri- Kohn-Wood, & Spencer, 2006), lower life satisfaction among vately identify as atheist do not publically identify as such for fear people of multiracial backgrounds (Giamo, Schmitt, & Outten, of ramifications (i.e., private vs. public self-identifications; Cragun 2012), and reduced self-esteem among homosexual and bisexual et al., 2012). Even so, it is clear that self-identified atheists are a men (Huebner, Rebchook, & Kegeles, 2004). Perceptions of dis- numerical minority in the United States today. crimination are also associated with lower physical well-being among African Americans (Borrell, Kiefe, Williams, Diez-Roux, The Marginalization of Atheism & Gordon-Larsen, 2006) and Mexican Americans (Finch, Hum- mer, Kolody, & Vega, 2001). In terms of the causal effect of Beyond being a numerical minority, atheists are also a margin- perceived discrimination on well-being, Schmitt, Branscombe, and alized group of people living in the United States (Gervais & Postmes (2003) found that manipulations of pervasive discrimina- Norenzayan, 2013). Several nationally representative surveys tion lowered self-esteem among women. In addition, Merritt, Ben- highlight the negative views that Americans express toward athe- nett, Williams, Edwards, and Sollers (2006) found that manipulat- ists. A 2007 Gallup poll, for example, assessed people’s willing- ing discrimination among African American men led to heightened ness to vote for hypothetical political candidates from different cardiovascular responses, which may negatively affect physical racial and religious backgrounds. Although many potential candi- well-being over time. In light of this research, we expected that dates would earn the majority of people’s votes, including an atheists’ perceptions of discrimination would be negatively asso- African American (94%), a woman (88%), and a homosexual ciated with their psychological and physical well-being. (55%), the only candidate that would not receive a majority vote was an atheist (45%; Jones, 2007). In a related survey, a majority The Rejection–Identification Model of people (53%) stated that they would be less likely to support a presidential candidate who was an atheist (Pew Research Center, Beyond documenting the relationship between discrimination 2014). This figure representing lack of support was larger than for and well-being, we also sought to test a process by which atheists all other characteristics, including a gay or lesbian candidate (27%) may protect themselves from the anticipated negative effects of and a candidate who had an extramarital affair (35%). As one more discrimination. As Allport (1954) argued, “misery finds balm example of people’s negative views toward atheists, 48% of people through the closer association of people who are miserable for the would disapprove of their children marrying an atheist (Edgell, same reason. Threats drive them to seek protective unity within Gerteis, & Hartmann, 2006). This disapproval rating was higher their common membership” (p. 148). Identification with a deval- This document is copyrighted by the American Psychological Association or one of its allied publishers. than ratings for all other minority groups included in the survey, ued group may be one of many ways that individuals respond to This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. such as Muslims (34%) and African Americans (19%). Edgell et stigma (Goffman, 1963). Similar to other marginalized groups, al. (2006) concluded that “atheists are less likely to be accepted, atheists may cope with discrimination by finding increased impor- publicly and privately, than any others from a long list of ethnic, tance in their stigmatized identity—a process of rejection and religious, and other minority groups” (p. 211). identification (Branscombe, Fernández, Gómez, & Cronin, 2011; Although these surveys indicate that Americans tend to hold Branscombe et al., 1999). The Rejection–Identification Model negative attitudes toward atheists, such studies do not speak to (i.e., RIM; Branscombe et al., 1999) posits that perceiving dis- whether atheists recognize their marginalized status. A few recent crimination based on one’s group membership may increase iden- studies, however, demonstrate that atheists do perceive themselves tification with that devalued group. Identification, in turn, reduces as targets of social rejection. For example, Cragun et al. (2012) the deleterious effects of discrimination on well-being (Haslam, found that approximately 41% of atheists experienced some form Jetten, Postmes, & Haslam, 2009). Rooted in Social Identity The- of discrimination during the five years before their study partici- ory (Tajfel, 1978; Tajfel & Turner, 1979), identification may pation. A related study found that discrimination toward atheists is provide people with a sense of belonging and perception of having pervasive, stemming from various sources (e.g., family members, supportive relationships with like-minded people, providing ben- DISCRIMINATION, ATHEIST IDENTIFICATION, & WELL-BEING 3

efits for well-being (Baumeister & Leary, 1995; Haslam et al., 2011). The conditions that atheists experience in the United States 2009). For example, people who strongly identify with atheism best reflect pervasive discrimination. The United States is a rela- may seek relationships with other atheists by means of becoming tively religious country where nearly 90% of people are fairly or involved in secular organizations (Ysseldyk, Matheson, & Anis- absolutely certain in the existence of god(s) and more than 80% of man, 2010). people affiliate with some religious group (Pew Research Center, Although the RIM focuses on social benefits of group identity, 2008). The religious context of this country is relevant as religion cognitive processes may also be protective. For example, identi- can act as a catalyst for intergroup hostility and intolerance (Sil- fication is associated with rejecting minority group stereotypes and berman, 2005), especially toward groups of people who are in resisting group-related stigma (Crabtree, Haslam, Postmes, & Has- violation of a given religion’s values (Shen, Yelderman, Haggard, lam, 2010). To the extent that discrimination threatens atheists’ & Rowatt, 2013; Whitley, 2009). This is important because athe- sense of certainty about their selves, their secular attitudes may ists have been found to be especially threatening to people’s values strengthen (McGregor, Zanna, Holmes, & Spencer, 2001), their (Cook, Cottrell, & Webster, 2014). Atheists also experience dis- group identification may increase (Grant & Hogg, 2012), and they crimination across many contexts, including slander by popular may further advocate for their beliefs (Gal & Rucker, 2010). As an media (e.g., TV and newspapers) and social condemnation by example, Farias, Newheiser, Kahane, and de Toledo (2013) found family members, friends, and coworkers (Hunsberger & Alte- that relatively secular individuals reported greater conviction in meyer, 2006; Hammer et al., 2012). Because the marginalization science when under stress and anxiety. In this way, identifying of atheism is widespread (Gervais & Norenzayan, 2013), cutting with one’s minority group and endorsing one’s minority beliefs in across personal and public domains of life (Edgell et al., 2006; the face of uncertainty (e.g., when being a target of discrimination) Hammer et al., 2012), atheists do appear to experience pervasive may be a mechanism to help bolster one’s sense of self. rather than infrequent discrimination in the United States. The process of rejection and identification has received em- Second, members of marginalized groups tend to increase their pirical support among many groups, including African Ameri- group identification when they perceive group differences as ille- cans (Branscombe et al., 1999), Mexican Americans (Romero & gitimate and when they believe that they can improve their status Roberts, 2003), international students (Schmitt, Spears, & by competing with the more dominant group (Branscombe et al., Branscombe, 2003), and older adults (Garstka, Schmitt, 2011; Ellemers, 1993; Tajfel & Turner, 1986). This second con- Branscombe, & Hummert, 2004). This process has also been dition is also satisfied for many atheists because they tend to documented in cross-sectional (Mossakowski, 2003), longitudinal perceive very distinct group boundaries between themselves and (Ramos, Cassidy, Reicher, & Haslam, 2012), and experimental religious individuals (Smith, 2011), and believe that the difference research (Jetten, Branscombe, Schmitt, & Spears, 2001). For ex- in social power across these two groups is illegitimate. The dif- ample, Jetten et al. (2001) manipulated perceptions of discrimina- ference in social power may cause some atheists to challenge and tion among people with body piercings. Participants in a high- compete with religion (Smith, 2013). By so doing, atheists main- discrimination condition reported stronger identification with other tain distinct group boundaries, thus staying closer to, rather than people who had body piercings compared with participants in a moving away from, their marginalized identity (Branscombe et al., low-discrimination condition. Higher levels of identification fur- 2011). The new atheist movement is an indication that some ther predicted greater self-esteem. Thus, there is evidence for a atheists join organizations of similar others to respond to perceived causal relationship of discrimination on identification, and for injustices against them by religious individuals and institutions identification’s benefits for well-being. Based on this previous (Guenther et al., 2013). research, we expected that perceiving discrimination would be Although atheists share experiences that are similar to those of associated with a higher degree of identification among atheists. In other marginalized groups included in previous RIM research, turn, identification as an atheist should reduce the negative effects there are at least two features of atheism that make the current of discrimination on well-being. There are alternative models, study a departure from most previous work: atheism is a conceal- however, that describe the relationship between discrimination and able identity, and atheism is belief-based. Previous studies testing well-being. For example, the discounting model (Crocker & Ma- the RIM have largely examined groups of people who have visible jor, 1989) predicts that discrimination should predict better well- stigmatized identities, such as group membership based on race, being under certain conditions. That is, being able to discount sex, and age. Atheists, however, have the option to “hide” their This document is copyrighted by the American Psychological Association or one of its allied publishers. negative treatment resulting from external causes (e.g., prejudice) identity from other people. Identification as an atheist is similar to This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. may be self-protective for some minority group members. The other identities that people can often keep concealed, including maladjustment model, as second alternative account, predicts that sexual orientation (Crocker, Major, & Steele, 1998; Pachankis, poor well-being may cause people to attribute negative treatment 2007; Quinn & Chaudoir, 2009). A recent study examined the to discrimination (Branscombe et al., 1999). We test for the plau- rejection–identification process for gay men and found that behav- sibility of these alternative models in the present study, comparing ioral identification in the form of involvement in the gay commu- results with the RIM. nity mediated the relationship between discrimination and self- esteem (Doyle & Molix, 2014). Behavioral identification implies a Rejection–Identification Model and Atheists degree of external visibility of one’s identity, such as endorsing magazines or TV shows directed toward the gay community. The We expect that the rejection–identification process is applicable definition of identification for the present study, however, focuses to atheists for at least two reasons. First, members of a marginal- on an internal evaluation of the importance and centrality of being ized group tend to increase their group identification when dis- an atheist. In addition to being a concealable identity, people crimination is pervasive rather than infrequent (Branscombe et al., actively choose to be an atheist by rejecting the notion of a god(s) 4 DOANE AND ELLIOTT

through a complex process of identity formation (Smith, 2011). AAI agreed to send e-mails inviting its members and supporters to Atheism is a voluntary or chosen identity to the extent that people participate.1 AAI also posted study information to its social media can decide their beliefs about religious matters. This type of websites, and encouraged its group members, or other secular identification is different from identities based on ascribed char- organizations, to distribute study information to their respective acteristics including race, sex, or other characteristics assigned memberships. E-mails and social media posts contained informa- from birth. In sum, the atheist identity differs from many other tion about this study along with a link to an online survey. identities examined with the RIM model because it is both con- The survey was active for six weeks, and 1,007 self-identified cealable and chosen. atheists completed survey items. Of these participants, the majority Despite the uniqueness of the atheist identity vis-a`-vis other (95%) indicated that they did not believe in the existence of god(s). groups for whom the RIM model applies, we argue that the process The other five percent of participants reported beliefs in god(s) that of rejection and identification would occur for atheists because did not align with theological atheism. For example, some self- their experiences in the United States meet two of the RIM’s identified atheists were not certain whether god(s) existed. We conditions highlighted earlier: the experience of pervasive discrim- excluded this subset of participants from all analyses so that our ination and the perception of group boundaries as being illegiti- sample included atheists as defined by both self-identifying and mate. Thus, there is reason to expect the concealable and voluntary theological atheism.2 Demographic characteristics of the final identity of being an atheist to function similarly to a visible and sample (n ϭ 960) are displayed in Table 1. ascribed identity in the context of the rejection–identification process. Identification as an atheist should become important in defining the self in the face of discrimination. Measures To test the RIM, we included measures of perceived discrimi- The Present Study nation, atheist identification, and well-being in the survey. Per- Although no known research has examined the intersection of ceived discrimination consisted of two related measures, including perceived discrimination, identification, and well-being among personal discrimination (i.e., personal experiences of discrimina- atheists, the rejection–identification process offers a rich theoret- tion) and group discrimination (i.e., perceptions of atheists as a ical model to guide this study. In sum, we expected that (a) derogated group). Atheist identification comprised the importance, perceived discrimination would be negatively associated with centrality, and commitment toward being an atheist. Well-being well-being, (b) perceived discrimination would be positively as- included measures of both psychological (i.e., self-esteem, life sociated with atheist identification, (c) atheist identification would satisfaction, and negative affect) and physical well-being. be positively associated with well-being, and (d) perceived dis- Personal discrimination. Five items assessed personal expe- crimination would be indirectly and positively associated with riences of discrimination. These items reflected common experi- well-being through atheist identification. ences of discrimination among atheists (see Hammer et al., 2012), The present study expands the net of previous RIM research in and were modeled from existing measures (e.g., Schmitt, at least three ways. First, this study tests the RIM’s predictions Branscombe, Kobrynowicz, & Owen, 2002; Branscombe et al., among atheists, an understudied group of people who face espe- 1999). An example item included I have felt isolated because I am cially negative attitudes from society at-large. Second, this study an atheist. Participants rated their agreement on a 5-point scale ϭ ϭ adds physical well-being as an independent outcome. Physical (1 strongly disagree to 5 strongly agree). well-being is a key outcome in studies that have assessed the Group discrimination. Six items assessed perceptions of consequences of discrimination (Pascoe & Smart Richman, 2009), atheists as a derogated group in the United States. Items were yet no known research has tested whether the rejection– similar to Schmitt et al.’s (2002) measure of out-group privilege, identification process applies to both psychological and physical and focused on how atheists fare compared with religious individ- well-being. Third, this study tests the RIM using structural equa- uals (mirroring the contrast between males and females in Schmitt tion modeling (SEM). Benefits of SEM include the ability to et al.’s work). An example item included Religious people have model first- and second-order latent factors with confirmatory more opportunities than do atheists in my country. Participants factor analysis, to account and correct for measurement error, and rated their agreement on a 5-point scale (1 ϭ strongly disagree to This document is copyrighted by the American Psychological Association or one of its alliedto publishers. model direct and indirect pathways (Henseler, 2012; Weston & 5 ϭ strongly agree). This article is intended solely for the personal use ofGore, the individual user and is not to be disseminated2006). broadly. Atheist identification. Three items assessed a related dimen- sion of identification as an atheist, including importance (i.e., Method Being an atheist is an important part of who I am), centrality (i.e., Being an atheist is not a central aspect of my identity; reverse coded), and commitment (i.e., I am committed to being an atheist). Participant Recruitment

Participants for this study included members of Atheist Alliance 1 Because of privacy concerns regarding AAI’s membership, no infor- International (AAI), an organization that supports education and mation is available to compute a response rate for this survey. Even if this activism addressing secular issues. AAI has different types of information were publically available, participants included people who members, including individual members (i.e., people who pay a may not have been AAI members, but instead received information through various social media posts. nominal membership fee) and group members (i.e., a collection of 2 Excluding the 47 participants who endorsed a belief in god(s), from secular organizations). There are also nonmember supporters, such uncertainty to certainty in the existence of god(s), does not change the as those who endorse AAI’s social media websites. For this study, pattern or significance of results. DISCRIMINATION, ATHEIST IDENTIFICATION, & WELL-BEING 5

Table 1 Results Demographic Characteristics

Percentage or Data Analysis Strategy Demographic characteristic mean (SD) We used Mplus 6 software (Muthén & Muthén, 2010) to test all Sex (1 ϭ female) 42% Age 43.76 (16.03) measurement and structural models. Measurement models describe Children (1 ϭ yes) 52% the associations between indicators (i.e., scale items) and their Education underlying constructs (i.e., latent factors). Structural models de- Less than secondary school 2% scribe the associations among factors (Weston & Gore, 2006). We Completed secondary school 16% Trade qualification 9% used multiple fit indices to assess the goodness of fit of both College/university degree 46% measurement and structural models, including the comparative fit Postgraduate degree 27% index (CFI), the root mean square error of approximation (RMSEA), Social class and the standardized root-mean-square residual (SRMR). Stan- Lower class 4% Working class 19% dards of good model fit are values at or above .90 for CFI, values Lower middle class 35% at or below .06 for RMSEA, and values at or below .08 for SRMR Upper middle class 40% (Hu & Bentler, 1995, 1999). We also report chi-square and relative Upper class 2% chi-square values (␹2/df; Wheaton, Muthén, Alwin, & Summers, Racial minority (1 ϭ racial minority) 8% Marital status (1 ϭ married) 66% 1977). As chi-square values tend to inflate with large sample sizes Employment status (1 ϭ employed) 56% (Schumacker & Lomax, 2004), relative chi-square values provide Region of residencea a better indicator of model fit. A liberal cutoff for a relative West 27% chi-square value is five or less (Wheaton et al., 1977), whereas a Midwest 33% South 27% conservative cutoff is two or less (Tabachnick & Fidell, 2007). Northeast 13% Setting of residence Rural 18% Confirmatory Factor Analyses Suburban 50% of Measurement Models Urban 32% Secular organization membership (1 ϭ member) 79% We first present confirmatory factor analyses (CFA) assessing Note. The structural model controls for the significant paths between model fit for first- and second-order latent factors. Table 2 presents factors and demographic characteristics. n ϭ 960. standardized factor loadings for first-order factors (all loadings at a Region of residence was dummy coded, with the Southern region serving p Ͻ .001). This second table also reports factor reliabilities that as the reference category. take into account the loadings and residual variances of their respective indicators (DeShon, 1998). Table 3 displays the means, standard deviations, and intercorrelations of the first-order factors. Participants rated their agreement on a 5-point scale (1 ϭ strongly Perceived discrimination. We estimated first-order latent disagree to 5 ϭ strongly agree). factors of personal and group discrimination. The first-order factor Self-esteem. The Rosenberg Self-Esteem Scale (Rosenberg, of personal discrimination fit the data well: ␹2(5) ϭ 16.60, p ϭ 1965) assessed personal self-esteem. This measure included 10 .005; ␹2/df ϭ 3.32; CFI ϭ .99; RMSEA ϭ .05; SRMR ϭ .02. The items (e.g., On the whole, I am satisfied with myself) rated on a first-order factor of group discrimination also fit the data well: 5-point scale (1 ϭ strongly disagree to 5 ϭ strongly agree). ␹2(9) ϭ 34.49, p Ͻ .001; ␹2/df ϭ 3.83; CFI ϭ .97; RMSEA ϭ .05; Life satisfaction. The Satisfaction with Life Scale (Diener, SRMR ϭ .03. These first-order factors had a strong and positive Emmons, Larsen, & Griffin, 1985) assessed judgments of overall correlation, r ϭ .63, p Ͻ .001. We treated these first-order factors life satisfaction. This measure contained five items (e.g., In most as “indicators” of perceived discrimination, the underlying second- ways, my life is close to my ideal) rated on a 7-point scale (1 ϭ order factor. This second-order factor fit the data well: ␹2(43) ϭ This document is copyrighted by the American Psychological Association or one of its alliedstrongly publishers. disagree to 7 ϭ strongly agree). 130.54, p Ͻ .001; ␹2/df ϭ 3.04; CFI ϭ .96; RMSEA ϭ .05; This article is intended solely for the personal use of the individual userNegative and is not to be disseminated broadly. affect. The Negative Affect Scale of the Positive SRMR ϭ .03. The standardized factor loadings were .72 (p Ͻ and Negative Affect Schedule Short Version (Watson, Clark, & .001) for personal discrimination and .88 (p Ͻ .001) for group Tellegen, 1988) assessed the degree to which participants experi- discrimination. enced five negative emotions. Participants indicated how often Atheist identification. We estimated the latent factor of athe- they felt upset, afraid, hostile, ashamed, and nervous during the ist identification, which included the three indicators of centrality, few weeks before their participation on a 5-point scale (1 ϭ very importance, and commitment. Because this factor consisted of slightly or not at all to 5 ϭ extremely). three indicators, no fit statistics for this CFA were calculated. Physical well-being. One item assessed physical well-being Psychological well-being. We estimated first-order latent fac- (i.e., In general, would you say your physical health is . . .) rated tors of life satisfaction, self-esteem, and negative affect. All factors on a 5-point scale (1 ϭ poor to 5 ϭ excellent). The subjective fit the data well, including life satisfaction, ␹2(3) ϭ 12.25, p ϭ .007; assessment of physical well-being reliably predicts objective mea- ␹2/df ϭ 4.08; CFI ϭ .99; RMSEA ϭ .06; SRMR ϭ .01, self-esteem, sures of health including mortality (Idler & Benyamini, 1997; Idler ␹2(31) ϭ 155.15, p Ͻ .001; ␹2/df ϭ 5; 4.08; CFI ϭ .97; RMSEA ϭ & Kasl, 1991). .07; SRMR ϭ .03, and negative affect, ␹2(3) ϭ 7.33, p ϭ .062; 6 DOANE AND ELLIOTT

Table 2 Standardized Factor Loadings of First-Order Factors

First-order factor Factor loadings

Personal discrimination (reliability ϭ .74)a 1. People generally think I am immoral because I am an atheist .64 2. I am treated no differently because I am an atheist Ϫ.63 3. I have felt isolated because I am an atheist .61 4. Being an atheist has not deprived me of any opportunities in my life Ϫ.55 5. I have personally experienced anti-atheist prejudice .58 Group discrimination (reliability ϭ .74)a 1. The media portray religious people as superior to atheists in my country .67 2. Anti-atheist prejudice is widespread in my country .55 3. Atheists are treated no differently than religious people in my country Ϫ.44 4. Religious people have more opportunities than do atheists in my country .63 5. Atheists are pressured to participate in religious activities in my country .54 6. Atheists have been the victims of hate crimes in my country .55 Atheist identification (reliability ϭ .68)a 1. Being an atheist is an important part of who I am .94 2. I am committed to being an atheist .41 3. Being an atheist is not a central aspect of my identity Ϫ.52 Life satisfaction (reliability ϭ .90)b 1. In most ways, my life is close to my ideal .92 2. The conditions of my life are excellent .79 3. I am satisfied with my life .81 4. So far I have gotten the important things I want in life .88 5. If I could live my life over, I would change almost nothing .58 Self-esteem (reliability ϭ .90)a 1. I take a positive attitude toward myself .87 2. On the whole, I am satisfied with myself .78 3. At times, I think I am no good at all Ϫ.74 4. I feel that I have a number of good qualities .49 5. I am able to do things as well as most other people .57 6. I feel I do not have much to proud of Ϫ.63 7. I certainly feel useless at times Ϫ.68 8. I feel that I am a person of worth, at least on an equal plane with others .66 9. I wish I could have more respect for myself Ϫ.68 10. All in all, I am inclined to feel that I am a failure Ϫ.77 Negative affect (reliability ϭ .76)c 1. Afraid .75 2. Upset .53 3. Hostile .35 4. Ashamed .67 5. Nervous .79 Note. All standardized factor loadings at p Ͻ .001. a Items rated on a 5-point scale (1 ϭ strongly disagree to 5 ϭ strongly agree). b Items rated on a 7-point scale (1 ϭ strongly disagree to 7 ϭ strongly agree). c Items rated on a 5-point scale (1 ϭ very slightly or not at all to 5 ϭ extremely).

␹2/df ϭ 2.44; CFI ϭ .99; RMSEA ϭ .04; SRMR ϭ .01.3 These crimination to atheist identification, psychological well-being, and

This document is copyrighted by the American Psychological Association or one of its alliedfactors publishers. were strongly correlated, including life satisfaction with physical well-being. We also specified paths from atheist identi- ϭ Ͻ ϭϪ Ͻ This article is intended solely for the personal use ofself-esteem, the individual user and is not to be disseminated broadly. r .74, p .001, and negative affect, r .51, p fication to both psychological and physical well-being. We al- .001, and self-esteem with negative affect, r ϭϪ.66, p Ͻ .001. We lowed residual variances to covary between psychological and phys- treated these first-order factors as “indicators” of psychological ical well-being. This final structural model fit the data well: ␹2(778) ϭ well-being, the underlying second-order factor. This second-order 1563.91, p Ͻ .001; ␹2/df ϭ 2.01; CFI ϭ .93; RMSEA ϭ .03; factor fit the data well: ␹2(159) ϭ 559.52, p Ͻ .001; ␹2/df ϭ 3.52; CFI ϭ .96; RMSEA ϭ .04; SRMR ϭ .05. The standardized factor loadings were .75 (p Ͻ .001) for life satisfaction, .98 (p Ͻ .001) for 3 We allowed residual variances to covary for several sets of indicators self-esteem, and Ϫ.68 (p Ͻ .001) for negative affect. within the first-order latent factors of psychological well-being, each of which was theoretically defensible (Hooper, Coughlan, & Mullen, 2008). For example, the residual variances for the indicators At times, I think I am Structural Model no good at all and I certainly feel useless at times of the first-order latent factor of self-esteem were allowed to correlate. These two indicators are After estimating and assessing model fit for first- and second- very similar and may measure a unique concept beyond self-esteem. The order latent factors, we estimated the final structural model to test correlated residual variances do not change the pattern or significance of the RIM. In this model, we specified paths from perceived dis- results. DISCRIMINATION, ATHEIST IDENTIFICATION, & WELL-BEING 7

Table 3 Means, Standard Deviations, and Bivariate Correlations Among First-Order Factors

Measure 1 2 3 4567

ءϪ.04 .09 ء08. 06. ءء24. ءءAtheist identification — .14 .1 ءءϪ.11 ءءء16. ءءءϪ.16 ءءءPersonal discrimination — .63 Ϫ.19 .2 ءϪ.07 ء09. ءءϪ.10 ءءGroup discrimination — Ϫ.12 .3 ءء38. ءءءϪ.51 ءءءLife satisfaction — .74 .4 ءءء39. ءءءSelf-esteem — Ϫ.66 .5 ءءءNegative affect — Ϫ.22 .6 7. Physical well-being — Mean 4.20 3.22 4.18 4.98 4.34 1.64 3.36 Standard deviation 0.78 0.88 0.63 1.40 0.70 0.60 0.99 Note. All measures represent first-order factors except for physical well-being, which is a one-item, manifest variable. .p Ͻ .001 ءءء .p Ͻ .01 ءء .p Ͻ .05 ء

SRMR ϭ .05. Figure 1 displays all standardized coefficients of the identification). Counter to the predictions of the discounting pathways in this model. model, perceived discrimination was a negative predictor of both Findings were entirely consistent with the rejection–identification psychological well-being (␤ϭϪ.22, p Ͻ .001) and physical process. First, perceived discrimination predicted both psycholog- well-being (␤ϭϪ.10, p ϭ .010). An additional model excluding ical well-being (␤ϭϪ.26, p Ͻ .001) and physical well-being physical well-being revealed the same pattern: perceived discrim- (␤ϭϪ.13, p ϭ .002). Participants who tended to experience ination is a negative predictor of psychological well-being discrimination and perceive atheists as a derogated group reported (␤ϭϪ.22, p Ͻ .001). lower psychological and physical well-being. Second, perceived We also tested a maladjustment model, which predicts that discrimination predicted atheist identification (␤ϭ.29, p Ͻ .001). people with relatively low well-being are more likely to perceive Aligning with the RIM, perceptions of discrimination may encour- a hostile environment characterized by discrimination directed age stronger identification as an atheist. Third, atheist identifica- toward them. This alternative model included pathways from psy- tion was a positive predictor of both psychological well-being chological and physical well-being to perceived discrimination. (␤ϭ.15, p Ͻ .001) and physical well-being (␤ϭ.12, p ϭ .002). Only psychological well-being was negatively associated with As expected, the two indirect pathways were also significant in perceived discrimination (␤ϭϪ.12, p ϭ .015). In contrast, this structural model. Perceived discrimination had indirect and physical well-being did not significantly predict perceived dis- positive effects via atheist identification on both psychological crimination (␤ϭϪ.05, p ϭ .241). Relative to these alternative well-being (␤ϭ.04, p ϭ .006) and physical well-being (␤ϭ.03, models, the RIM receives the most support in validating the p ϭ .012). According to the process of rejection and identification, predicted relationships among perceived discrimination, atheist perceiving discrimination may encourage identification as an athe- identification, and psychological and physical well-being. ist, which in turn appears to convey benefits for well-being. Given that the direct effects of perceived discrimination on psychological and physical well-being were larger than their total effects Discussion (␤ϭϪ.26 vs. ␤ϭϪ.21, and ␤ϭϪ.13 vs. ␤ϭϪ.10, respec- Perceiving discrimination based on one’s group membership is tively), and the total effects were statistically significant, atheist generally associated with less favorable psychological and physi- identification appears to be helpful in reducing, but not eliminat- cal well-being (Pascoe & Smart Richman, 2009). Both early (e.g., ing, discrimination’s ill effects on well-being. Findings are ad- Allport, 1954) and recent researchers (e.g., Branscombe et al., justed for significant effects of several demographic characteris- 1999) have argued that perceptions of discrimination can increase 4

This document is copyrighted by the American Psychological Association or one of its alliedtics. publishers. identification with one’s marginalized group. Under certain cir-

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. cumstances, increased identification confers benefits for the indi- Alternative Models vidual. According to the rejection–identification process (Branscombe In addition to testing the RIM, we tested alternative models. Most relevant to this study is Crocker and Major’s (1989) dis- 4 The only significant demographic characteristic that predicted per- counting model, which posits that perceptions of discrimination ceived discrimination was age (␤ϭϪ.18, p Ͻ .001). Significant predictors ␤ϭ Ͻ may be associated with better, rather than worse well-being under of atheist identification were age ( .17, p .001) and setting of residence (ranged from rural to urban; ␤ϭ.09, p ϭ .009). Significant certain circumstances. According to this model, individuals who predictors of psychological well-being were setting of residence (ranged make an attribution to discrimination may protect overall well- from rural to urban; ␤ϭϪ.08, p ϭ .010), having children (vs. not having being by shifting the cause of rejection from the self to external children; ␤ϭ.21, p Ͻ .001), being married (vs. not being married; ␤ϭ discriminatory practices (Crocker & Major, 1989). We tested a .13, p ϭ .001), and being a member of a secular organization (vs. being a supporter; ␤ϭ.07, p ϭ .028). Significant predictors of physical well-being model in which discrimination predicted psychological and phys- were setting of residence (ranged from rural to urban; ␤ϭϪ.09, p ϭ .004), ical well-being (excluding atheist identification since the discount- education (␤ϭ.14, p Ͻ .001), and being employed (vs. not employed; ␤ϭ ing model does not make any specific predictions for the role of .11, p Ͻ .001). 8 DOANE AND ELLIOTT

Self-Esteem .93***

.79*** Psychological Life Well-Being Satisfaction -.69 *** Negative -.26*** Affect (-.21***) e .15*** Personal Discrimination .75***

Perceived .29*** Atheist Discrimination Identification .37*** Group .83*** Discrimination .12** -.13** (-.10*)

Physical Well-Being e

Figure 1. Structural model of the rejection–identification process. ␹2(778) ϭ 1563.91, p Ͻ .001; ␹2/df ϭ 2.01; CFI ϭ .93; RMSEA ϭ .03; SRMR ϭ .05. This figure displays standardized regression coefficients. Total effects .p Ͻ .001 ءءء ,p Ͻ .01 ءء ,p Ͻ .05 ء .of perceived discrimination on well-being are in parentheses. n ϭ 960

et al., 1999), identification weakens the negative effects of discrim- example, Jetten et al. (2001) demonstrated that manipulating per- ination on well-being. This model has received empirical support ceptions of discrimination led to increased identification among among various marginalized groups, including African Americans people with body piercings. Relevant to atheists, Ysseldyk, Has- (Branscombe et al., 1999), Mexican Americans (Romero & Rob- lam, Matheson, and Anisman (2012) found that manipulating erts, 2003), and Filipino Americans (Mossakowski, 2003). The feelings of threat increased atheists’ own positive feelings toward current study adds to this previous research by finding support of other atheists. In addition, Merritt et al. (2006) found that the the RIM among a sample of atheists living in the United States. manipulation of discrimination heightened cardiovascular re- Atheists’ perceptions of discrimination (i.e., personal experiences sponses among African American men. In light of these various of discrimination and perceptions of atheists as a derogated group) findings, it is clear that the relationships among discrimination, were associated with lower psychological and physical well-being. identity, and well-being are complex. Though there is evidence of While having a direct negative effect on well-being, perceptions of a causal link of perceived discrimination on identification, for discrimination also had an indirect and positive effect on well- example, their effects are likely reciprocal. As Cragun et al. (2012) being through increased identification as an atheist. In the face of suggest, highly identified atheists are likely to experience more social rejection, atheists may be resilient by drawing benefits from discrimination because of the heightened attention an atheist iden- their identification as an atheist. To our knowledge, this is the first tification yields from other people. Highly identified atheists are test of the rejection–identification process among atheists. also likely to find themselves in situations in which they experi- Although findings from this study add to current literature on ence more discrimination, such as when protesting. To ensure that the rejection–identification process, a few limitations warrant fur- similar causal patterns emerge among atheists as posited by the ther discussion. First, the data are cross-sectional, so we cannot RIM, future research should manipulate perceptions of discrimi- conclusively claim any causal links between the paths outlined in nation to measure its causal effect on both atheist identification the RIM, including perceptions of discrimination affecting atheist and well-being. identification and well-being. There are alternative accounts for Second, this study examined the rejection–identification process This document is copyrighted by the American Psychological Association or one of its allied publishers. the link between discrimination and identification. For example, among many “card-carrying atheists,” or people who both self- This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. group identification plays a role in how members of marginalized identify as atheist and join or support organizations catering to groups experience discrimination and attribute events to discrim- nonreligious people (Caldwell-Harris, Wilson, LoTempio, & Beit- ination (Eccleston & Major, 2006). In one longitudinal study Hallahmi, 2011). Atheists may join or subscribe to online organi- among African Americans, Sellers and Shelton (2003) found that zations for various reasons, such as perceiving online spaces as the centrality of participants’ racial identity predicted future expe- places to “come out, speak out, and ‘meet up’” (Cimino & Smith, riences of discrimination after controlling for initial experiences of 2011, p. 31). Atheists who strongly identify as such are probably discrimination. There is also experimental evidence demonstrating more likely to join organizations relative to their less identified that highly identified members of marginalized groups (e.g., counterparts. Descriptive analyses suggest that participants in our women) are more likely to attribute ambiguous and negative sample were indeed highly identified as being atheist, scoring an events to discrimination compared to less identified members average of 4.20 (SD ϭ 0.78) on a 5-point scale of atheist identi- (Major, Quinton, & Schmader, 2003). fication. Thus, our results can only speak to atheists who are Previous research, however, has tested for a causal link of (a) highly identified, and (b) seek organizational involvement. It is discrimination on group identification, as posited by the RIM. For possible that atheists who are more isolated (i.e., without organi- DISCRIMINATION, ATHEIST IDENTIFICATION, & WELL-BEING 9

zational affiliation) might suffer more from their stigmatized sta- such studies. The general exclusion of the nonreligious has been a tus, and identification might do little, if anything, to offset stigma’s criticism of this line of research (see Hwang, Hammer, & Cragun, ill effects. That is, atheists who do not join organizations might 2011). However, there are a few recent and notable examples of lack the apparent benefits of an atheist identity as a result, at least research that addresses such criticism, and these should provide a in part, of not having a social network of like-minded others. model for future studies to follow. For example, in a comparison Alternatively, some atheists who are less engaged in organized across 64 countries, Stavrova, Fetchenhauer, and Schlösser (2013) atheist activities might use more individualist coping strategies to found that the gap in happiness and life satisfaction between the deal with social rejection, including keeping their identification as religious and nonreligious is wider in countries where religiosity is an atheist concealed from others. Concealing this marginalized more normative (Study 1). These researchers further demonstrated identity may reduce the direct experiences of discrimination that religious individuals living in countries that are more religious among such isolated atheists. Because atheists are a hard-to-reach perceive greater social respect and recognition relative to their population of people, soliciting participants via online sources is nonreligious counterparts, partially accounting for their gap in convenient and an appropriate first step, yet future studies should well-being (Study 2). Similarly, Diener, Tay, and Myers (2011) aim to reach a broader selection of atheists when possible. found that social support, perceived social respect, and sense of Third, the present study assessed identification as a composite of purpose or meaning mediated the relationship between religiosity importance, centrality, and commitment, yet identification in- and indicators of subjective well-being (Study 2). Further, the gap cludes other dimensions. For example, Giamo et al., (2012) tested in subjective well-being between the more and less religious was the RIM with a comprehensive measure of multiracial identifica- larger in relatively religious countries compared with less reli- tion based on Leach et al.’s (2008) five dimensions of group gious countries. Other studies have demonstrated a similar effect identification: in-group solidarity, identity centrality, self- whereby the benefits of religion for well-being are context-depen- stereotyping, identity satisfaction, and in-group homogeneity. By dent: religion is beneficial in countries that are more religious but using this multidimensional approach, Giamo et al. (2012) exam- not necessarily in countries that are less religious (see Gebauer, ined the specific aspects of multiracial identification that uniquely Sedikides, & Neberich, 2012; Okulicz-Kozaryn, 2010). Research- mediated the effect of perceived discrimination on well-being. ers should aim to document other moderators of the link between Among multiracial participants, self-stereotyping (i.e., extent to religion and well-being, such as the degree of government restric- which people see themselves as similar to an average member of tions on individual liberties (see Elliott & Hayward, 2009; Hay- the in-group) was the only aspect of identification to emerge as a ward & Elliott, 2014). significant mediator within the context of the RIM. The measure of Second, because being an atheist is associated with stigma, an identification in the present study closely resembles that of identity important social issue emerges regarding how the United States centrality, or the importance of one’s identification as an atheist. It and other countries can actually reduce stigma toward atheists. A is unclear whether identity centrality alone explains our findings, recent line of research points to the effects of atheist prevalence or whether one or more aspects of identification (e.g., in-group (Gervais, 2011) and secular authority (Gervais & Norenzayan, solidarity or identity satisfaction) would prove useful in explaining 2012) on attenuating antiatheist prejudice. For example, religious the indirect path predicted by the RIM. Future research is neces- individuals in countries with larger numbers of atheists report less sary to determine the unique effects of different dimensions of negative attitudes toward atheists compared with religious people identification on well-being among atheists specifically, and peo- in countries with fewer atheists, and people’s perceptions of the ple of other marginalized groups more broadly. prevalence of atheists has a causal impact on reducing antiatheist Fourth, whereas the present study modeled a direct link between prejudice (Gervais, 2011). In addition, increasing people’s aware- atheist identification and well-being, future avenues of research ness of secular authority decreases distrust toward atheists (Ger- should explore the mechanisms through which identification con- vais & Norenzayan, 2012). Given that we have demonstrated that veys its benefits (see Crabtree et al., 2010). For example, identi- discrimination is associated with lower well-being among atheists, fication may provide individuals with opportunities for developing future research should continue to examine ways to reduce the social relationships and enhance perceptions of social support negative attitudes that people harbor toward atheists. (Schmitt & Branscombe, 2002). Identification also provides peo- ple with a sense of purpose, belonging, and acceptance, which has This document is copyrighted by the American Psychological Association or one of its allied publishers. Conclusion the potential to promote well-being (Baumeister & Leary, 1995; This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Haslam et al., 2009). The benefits of identification work through There is very little research focusing on atheists, including cognitive routes as well, including the rejection of minority group examining the consequences of marginalization for these people stereotypes (Crabtree et al., 2010). The benefits of atheist identi- (Brewster, Robinson, Sandil, Esposito, & Geiger, 2014). This is fication on well-being are indirect, yet research needs to distin- surprising as atheists continue to rank among the least accepted guish the mechanisms that explain social identification’s benefits groups of people in the United States. Like people who belong to for both psychological and physical well-being. other marginalized groups, perceptions of discrimination predict poor psychological and physical well-being among atheists. One Future Directions way that atheists may cope with such discrimination is by further believing that being an atheist is important and central to their There are at least two broad research aims related to the present identity. This rejection and identification process may ultimately topic that should guide future research. First, although there has protect well-being by reducing discrimination’s negative effects. It been substantial interest in studying the possible role of religion in is our hope that this study will provide the impetus to document the promoting well-being, nonreligious people are rarely included in potential consequences of social rejection for atheists. As research- 10 DOANE AND ELLIOTT

ers have noted (e.g., Norenzayan & Gervais, 2013), many nations Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. Gilbert, around the world are experiencing a transition from religious S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology (4th fervor to religious disinterest and even disbelief. It is increasingly ed., pp. 504–553). Boston, MA: McGraw-Hill. becoming important to understand further the atheist experience, Dawkins, R. (2006). The God delusion. Boston, MA: Mariner. from the development of atheist identification and its impact on DeShon, R. P. (1998). A cautionary note on measurement error corrections well-being to intergroup relations between atheists and their reli- in structural equation models. Psychological Methods, 3, 412–423. gious counterparts. doi:10.1037/1082-989X.3.4.412 Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction with Life Scale. Journal of Personality Assessment, 49, References 71–75. doi:10.1207/s15327752jpa4901_13 Allport, G. W. (1954). The nature of prejudice. Reading, MA: Addison Diener, E., Tay, L., & Myers, D. G. (2011). The religion paradox: If Wesley. religion makes people happy, why are so many dropping out? Journal of Banks, K. H., Kohn-Wood, L. P., & Spencer, M. (2006). An examination Personality and Social Psychology, 101, 1278–1290. doi:10.1037/ of the African American experience of everyday discrimination and a0024402 psychological distress. Community Mental Health Journal, 42, 555–570. Doyle, D. M., & Molix, L. (2014). Perceived discrimination and well-being doi:10.1007/s10597-006-9052-9 in gay men: The protective role of behavioural identification. Psychol- Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for ogy & Sexuality, 5, 117–130. doi:10.1080/19419899.2011.653689 interpersonal attachments as a fundamental human motivation. Psycho- Dubow, E. F., Pargament, K. I., Boxer, P., & Tarakeshwar, N. (2000). logical Bulletin, 117, 497–529. doi:10.1037/0033-2909.117.3.497 Initial investigation of Jewish early adolescents’ ethnic identity, stress, Borrell, L. N., Kiefe, C. I., Williams, D. R., Diez-Roux, A. V., & Gordon- and coping. The Journal of Early Adolescence, 20, 418–441. doi: Larsen, P. (2006). Self-reported health, perceived racial discrimination, 10.1177/0272431600020004003 and skin color in African Americans in the CARDIA Study. Social Eccleston, C. P., & Major, B. N. (2006). Attributions to discrimination and Science & Medicine, 63, 1415–1427. doi:10.1016/j.socscimed.2006.04 self-esteem: The role of group identification and appraisals. Group .008 Processes & Intergroup Relations, 9, 147–162. doi:10.1177/ Branscombe, N. R., Fernández, S., Gómez, A., & Cronin, T. (2011). 1368430206062074 Moving toward or away from a group identity: Different strategies for Edgell, P., Gerteis, J., & Hartmann, D. (2006). Atheists as ‘other’: Moral coping with pervasive discrimination. In J. Jetten, C. Haslam, & S. boundaries and cultural membership in American society. American Haslam (Eds.), The social cure: Identity, health and well-being (pp. Sociological Review, 71, 211–234. doi:10.1177/000312240607100203 115–131). New York, NY: Psychology Press. Ellemers, N. (1993). The influence of socio-structural variables on identity Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Perceiving management strategies. European Review of Social Psychology, 4, 27– pervasive discrimination among African Americans: Implications for 57. doi:10.1080/14792779343000013 group identification and well-being. Journal of Personality and Social Psychology, 77, 135–149. doi:10.1037/0022-3514.77.1.135 Elliott, M., & Hayward, R. D. (2009). Religion and life satisfaction Brewster, M. E., Robinson, M. A., Sandil, R., Esposito, J., & Geiger, E. worldwide: The role of government regulation. Sociology of Religion, (2014). Arrantly absent: Atheism in psychological science from 2001 to 70, 285–310. doi:10.1093/socrel/srp028 2012. The Counseling Psychologist, 42, 628–663. doi:10.1177/ Farias, M., Newheiser, A. K., Kahane, G., & de Toledo, Z. (2013). 0011000014528051 Scientific faith: Belief in science increases in the face of stress and Caldwell-Harris, C. L., Wilson, A. L., LoTempio, E., & Beit-Hallahmi, B. existential anxiety. Journal of Experimental Social Psychology, 49, (2011). Exploring the atheist personality: Well-being, awe, and magical 1210–1213. doi:10.1016/j.jesp.2013.05.008 thinking in atheists, Buddhists, and Christians. Mental Health, Religion Finch, B. K., Hummer, R. A., Kolody, B., & Vega, W. A. (2001). The role & Culture, 14, 659–672. doi:10.1080/13674676.2010.509847 of discrimination and acculturative stress in Mexican-origin adults’ Cimino, R., & Smith, C. (2011). The new atheism and the formation of the physical health. Hispanic Journal of Behavioral Sciences, 23, 399–429. imagined secularist community. Journal of Media and Religion, 10, doi:10.1177/0739986301234004 24–38. doi:10.1080/15348423.2011.549391 Franks, A. S. and Scherr, K. C. (in press). A sociofunctional approach to Clark, R., Anderson, N., Clark, V., & Williams, D. (1999). Racism as a prejudice at the polls: Are atheists more politically disadvantaged than stressor for African Americans. American Psychologist, 54, 805–816. gays and blacks? Journal of Applied Social Psychology. doi:10.1037/0003-066X.54.10.805 Gal, D., & Rucker, D. D. (2010). When in doubt, shout!: Paradoxical Cliteur, P. (2009). The definition of atheism. Journal of Religion and influences of doubt on proselytizing. Psychological Science, 21, 1701– This document is copyrighted by the American Psychological Association or one of its allied publishers. Society, 11, 1–23. 1707. doi:10.1177/0956797610385953

This article is intended solely for the personal use ofCook, the individual user and is not to beC. disseminated broadly. L., Cottrell, C. A., & Webster, G. D. (2014). No good without Garstka, T. A., Schmitt, M. T., Branscombe, N. R., & Hummert, M. (2004). God: Antiatheist prejudice as a function of threats to morals and values. How young and older adults differ in their responses to perceived age Psychology of Religion and Spirituality. Advance online publication. discrimination. Psychology and Aging, 19, 326–335. doi:10.1037/0882- doi:10.1037/rel0000013 7974.19.2.326 Crabtree, J. W., Haslam, S., Postmes, T., & Haslam, C. (2010). Mental Gebauer, J. E., Sedikides, C., & Neberich, W. (2012). Religiosity, social health support groups, stigma, and self-esteem: Positive and negative implications of group identification. Journal of Social Issues, 66, 553– self-esteem, and psychological adjustment: On the cross-cultural speci- 569. doi:10.1111/j.1540-4560.2010.01662.x ficity of the psychological benefits of religiosity. Psychological Science, Cragun, R. T., Kosmin, B., Keysar, A., Hammer, J. H., & Nielsen, M. 23, 158–160. doi:10.1177/0956797611427045 (2012). On the receiving end: Discrimination toward the non-religious in Gervais, W. M. (2011). Finding the faithless: Perceived atheist prevalence the United States. Journal of Contemporary Religion, 27, 105–127. reduces anti-atheist prejudice. Personality and Social Psychology Bul- doi:10.1080/13537903.2012.642741 letin, 37, 543–556. doi:10.1177/0146167211399583 Crocker, J., & Major, B. (1989). Social stigma and self-esteem: The Gervais, W. M. (2013). In Godlessness we distrust: Using social psychol- self-protective properties of stigma. Psychological Review, 96, 608– ogy to solve the puzzle of anti-atheist prejudice. Social and Personality 630. doi:10.1037/0033-295X.96.4.608 Psychology Compass, 7(6), 366–377. doi:10.1111/spc3.12035 DISCRIMINATION, ATHEIST IDENTIFICATION, & WELL-BEING 11

Gervais, W. M., & Norenzayan, A. (2012). Reminders of secular authority Hwang, K., Hammer, J. H., & Cragun, R. T. (2011). Extending religion- reduce believers’ distrust of atheists. Psychological Science, 23, 483– health research to secular minorities: Issues and concerns. Journal of 491. doi:10.1177/0956797611429711 Religion and Health, 50, 608–622. doi:10.1007/s10943-009-9296-0 Gervais, W. M., & Norenzayan, A. (2013). Religion and the origins of Idler, E. L., & Benyamini, Y. (1997). Self-rated health and mortality: A anti-atheist prejudice. In S. Clarke, R. Powell, & J. Savulescu (Eds.), review of twenty-seven community studies. Journal of Health and Intolerance and conflict: A scientific and conceptual investigation (pp. Social Behavior, 38, 21–37. doi:10.2307/2955359 126–145). Oxford, UK: Oxford University Press. doi:10.1093/acprof: Idler, E. L., & Kasl, S. (1991). Health perceptions and survival: Do global oso/9780199640911.003.0007 evaluations of health status really predict mortality? Journal of Geron- Gervais, W. M., Shariff, A. F., & Norenzayan, A. (2011). Do you believe tology, 46, S55–S65. doi:10.1093/geronj/46.2.S55 in atheists? Distrust is central to anti-atheist prejudice. Journal of Per- Jetten, J., Branscombe, N. R., Schmitt, M. T., & Spears, R. (2001). Rebels sonality and Social Psychology, 101, 1189–1206. doi:10.1037/a0025882 with a cause: Group identification as a response to perceived discrimi- Giamo, L. S., Schmitt, M. T., & Outten, H. (2012). Perceived discrimina- nation from the mainstream. Personality and Social Psychology Bulletin, tion, group identification, and life satisfaction among multiracial people: 27, 1204–1213. doi:10.1177/0146167201279012 A test of the rejection–identification model. Cultural Diversity & Ethnic Johnson, M. K., Rowatt, W. C., & LaBouff, J. P. (2012). Religiosity and Minority Psychology, 18, 319–328. doi:10.1037/a0029729 prejudice revisited In-group favoritism, out-group derogation, or both? Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Psychology of Religion and Spirituality, 4, 154–168. doi:10.1037/ New York, NY: Simon & Schuster. a0025107 Grant, F., & Hogg, M. A. (2012). Self-uncertainty, social identity promi- Jones, J. M. (2007). Some Americans reluctant to vote for Mormon, nence and group identification. Journal of Experimental Social Psychol- 72-year-old presidential candidates. Gallup News Service. Retrieved ogy, 48, 538–542. doi:10.1016/j.jesp.2011.11.006 from http://www.gallup.com/poll/26611/some-americans-reluctant- Guenther, K. M., Mulligan, K., & Papp, C. (2013). From the outside in: vote-mormon-72yearold-presidential-candidates.aspx Crossing boundaries to build collective identity in the new atheist Keysar, A. (2007). Who are America’s atheists and agnostics? In B. A. movement. Social Problems, 60, 457–475. doi:10.1525/sp.2013.60.4 Kosmin & A. Keysar (Eds.), Secularism and secularity: Contemporary .457 international perspectives (pp. 33–39). Hartford, CT: Institute for the Hammer, J. H., Cragun, R. T., Hwang, K., & Smith, J. M. (2012). Forms, Study of Secularism in Society and Culture. frequency, and correlates of perceived anti-atheist discrimination. Sec- Kosmin, B. A., & Keysar, A. (2006). Religion in a free market: Religion ularism and Nonreligion, 1, 43–67. doi:10.5334/snr.ad and non-religious Americans. Ithaca, NY: Paramount Market Publish- Harper, M. (2007). The stereotyping of nonreligious people by religious ing. students: Contents and subtypes. Journal for the Scientific Study of Kosmin, B. A., & Keysar, A. (2009). American Religious Identification Religion, 46, 539–552. doi:10.1111/j.1468-5906.2007.00376.x Survey (ARIS) 2008. Hartford, CT: Trinity College. Retrieved from Haslam, S. A., Jetten, J., Postmes, T., & Haslam, C. (2009). Social identity, http://commons.trincoll.edu/aris/files/2011/08/ARIS_Report_2008.pdf health and well-being: An emerging agenda for applied psychology. Leach, C. W., van Zomeren, M., Zebel, S., Vliek, M. W., Pennekamp, Applied Psychology: An International Review, 58, 1–23. doi:10.1111/j S. F., Doosje, B.,...Spears, R. (2008). Group-level self-definition and .1464-0597.2008.00379.x self-investment: A hierarchical (multicomponent) model of in-group Hayward, R. D., & Elliott, M. (2014). Cross-national analysis of the identification. Journal of Personality and Social Psychology, 95, 144– influence of cultural norms and government restrictions on the relation- 165. doi:10.1037/0022-3514.95.1.144 ship between religion and well-being. Review of Religious Research, 56, Major, B., Quinton, W. J., & Schmader, T. (2003). Attributions to discrim- 23–43. doi:10.1007/s13644-013-0135-0 ination and self-esteem: Impact of group identification and situational Henseler, J. (2012). Why generalized structured component analysis is not Journal of Experimental Social Psychology, 39, universally preferable to structural equation modeling. Journal of the ambiguity. 220–231. Academy of Marketing Science, 40, 402–413. doi:10.1007/s11747-011- doi:10.1016/S0022-1031(02)00547-4 0298-6 McGregor, I., Zanna, M. P., Holmes, J. G., & Spencer, S. J. (2001). Hitchens, C. (2007). God is not great: How religion poisons everything. Compensatory conviction in the face of personal uncertainty: Going to New York, NY: Twelve. extremes and being oneself. Journal of Personality and Social Psychol- Hooper, D., Coughlan, J., & Mullen, M. R. (2008). Structural equation ogy, 80, 472–488. doi:10.1037/0022-3514.80.3.472 modelling: Guidelines for determining model fit. The Electronic Journal Merritt, M. M., Bennett, G., Jr., Williams, R. B., Edwards, C. L., & Sollers, of Business Research Methods, 6, 53–60. J. (2006). Perceived racism and cardiovascular reactivity and recovery to Hout, M., & Fischer, C. S. (2002). Why more Americans have no religious personally relevant stress. Health Psychology, 25, 364–369. doi: This document is copyrighted by the American Psychological Association or one of its allied publishers. preference: Politics and generations. American Sociological Review, 67, 10.1037/0278-6133.25.3.364 This article is intended solely for the personal use of the individual user165–190. and is not to be disseminated broadly. doi:10.2307/3088891 Mossakowski, K. N. (2003). Coping with perceived discrimination: Does Hu, L. T., & Bentler, P. M. (1995). Evaluating model fit. In R. H. Hoyle ethnic identity protect mental health? Journal of Health and Social (Ed.), Structural equation modeling: Concepts, issues and applications Behavior, 44, 318–331. doi:10.2307/1519782 (pp. 76–99). Thousand Oaks, CA: Sage. Muthén, L. K., & Muthén, B. O. (2010). Mplus user’s guide (6th ed.). Los Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in Angeles, CA: Author. covariance structure analysis: Conventional criteria versus new alterna- Norenzayan, A., & Gervais, W. M. (2013). The origins of religious tives. Structural Equation Modeling, 6, 1–55. doi:10.1080/ disbelief. Trends in Cognitive Sciences, 17, 20–25. doi:10.1016/j.tics 10705519909540118 .2012.11.006 Huebner, D. M., Rebchook, G. M., & Kegeles, S. M. (2004). Experiences Okulicz-Kozaryn, A. (2010). Religiosity and life satisfaction across na- of harassment, discrimination, and physical violence among young gay tions. Mental Health, Religion & Culture, 13, 155–169. doi:10.1080/ and bisexual men. American Journal of Public Health, 94, 1200–1203. 13674670903273801 doi:10.2105/AJPH.94.7.1200 Pachankis, J. E. (2007). The psychological implications of concealing a Hunsberger, B. E., & Altemeyer, B. (2006). Atheists: A groundbreaking stigma: A cognitive-affective-behavioral model. Psychological Bulletin, study of America’s nonbelievers. Amherst, NY: Prometheus Books. 133, 328–345. doi:10.1037/0033-2909.133.2.328 12 DOANE AND ELLIOTT

Pascoe, E. A., & Smart Richman, L. (2009). Perceived discrimination and system analysis. In R. F. Paloutzian & C. L. Park (Eds.), Handbook of health: A meta-analytic review. Psychological Bulletin, 135, 531–554. religion and spirituality (pp. 529–549). New York, NY: Guilford Press. doi:10.1037/a0016059 Smith, C., & Cimino, R. (2012). Atheisms unbound: The role of the new Pew Research Center. (2008). U.S. religious landscape survey. Retrieved media in the formation of a secularist identity. Secularism and Nonre- from http://religions.pewforum.org/pdf/report-religious-landscape- ligion, 1, 17–31. doi:10.5334/snr.ab study-full.pdf Smith, J. M. (2011). Becoming an atheist in America: Constructing identity Pew Research Center. (2012). “Nones” on the rise: One-in-five adults have and meaning from the rejection of theism. Sociology of Religion, 72, no religious affiliation. Retrieved from http://www.pewforum.org/ 215–237. doi:10.1093/socrel/srq082 uploadedFiles/Topics/Religious_Affiliation/Unaffiliated/NonesOnThe Smith, J. M. (2013). Creating a godless community: The collective identity Rise-full.pdf work of contemporary American atheists. Journal for the Scientific Pew Research Center. (2014). For 2016 hopefuls, WA experience could do Study of Religion, 52, 80–99. doi:10.1111/jssr.12009 more harm than good. Retrieved from http://www.people-press.org/files/ Stavrova, O., Fetchenhauer, D., & Schlösser, T. (2013). Why are religious legacy-pdf/5-19-14%20Presidential%20Traits%20Release.pdf people happy? The effect of the social norm of religiosity across coun- Quinn, D. M., & Chaudoir, S. R. (2009). Living with a concealable tries. Social Science Research, 42, 90–105. doi:10.1016/j.ssresearch stigmatized identity: The impact of anticipated stigma, centrality, sa- .2012.07.002 lience, and cultural stigma on psychological distress and health. Journal Streib, H., & Klein, C. (2013). Atheists, agnostics, and apostates. In K. I. of Personality and Social Psychology, 97, 634–651. doi:10.1037/ Pargament, J. J. Exline, & J. W. Jones (Eds.), APA handbook of psy- a0015815 chology, religion, and spirituality: Context, theory, and research (Vol. Ramos, M. R., Cassidy, C., Reicher, S., & Haslam, S. (2012). A longitu- 1, pp. 713–728). Washington, DC: American Psychological Association. dinal investigation of the rejection–identification hypothesis. British Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th Journal of Social Psychology, 51, 642–660. doi:10.1111/j.2044-8309 ed.). New York, NY: Allyn & Bacon. .2011.02029.x Tajfel, H. (1978). The social psychology of minorities. London, UK: Ritter, R. S., & Preston, J. L. (2011). Gross gods and icky atheism: Disgust Minority Rights Group. responses to rejected religious beliefs. Journal of Experimental Social Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup Psychology, 47, 1225–1230. doi:10.1016/j.jesp.2011.05.006 conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of Romero, A. J., & Roberts, R. E. (2003). The impact of multiple dimensions intergroup relations (pp. 33–48). Pacific Grove, CA: Brooks/Cole. of ethnic identity on discrimination and adolescents’ self-esteem. Jour- Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup nal of Applied Social Psychology, 33, 2288–2305. doi:10.1111/j.1559- behavior. In S. Worchel & W. Austin (Eds.), Psychology of intergroup 1816.2003.tb01885.x relations (pp. 7–24). Chicago, IL: Nelson-Hall. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and vali- NJ: Princeton University Press. dation of brief measures of positive and negative affect: The PANAS Schmitt, M. T., & Branscombe, N. R. (2002). The meaning and conse- scales. Journal of Personality and Social Psychology, 54, 1063–1070. quences of perceived discrimination in disadvantaged and privileged doi:10.1037/0022-3514.54.6.1063 social groups. In W. Stroebe & M. Hewstone (Eds.), European review of Weber, S. R., Pargament, K. I., Kunik, M. E., Lomax, J., & Stanley, M. A. social psychology (Vol. 12, pp. 167–199). Chichester, UK: Wiley. (2012). Psychological distress among religious nonbelievers: A system- doi:10.1080/14792772143000058 atic review. Journal of Religion and Health, 51, 72–86. doi:10.1007/ Schmitt, M. T., Branscombe, N. R., Kobrynowicz, D., & Owen, S. (2002). s10943-011-9541-1 Perceiving discrimination against one’s gender group has different im- Weston, R., & Gore, P. A., Jr. (2006). A brief guide to structural equation plications for well-being in women and men. Personality and Social modeling. The Counseling Psychologist, 34, 719–751. doi:10.1177/ Psychology Bulletin, 28, 197–210. doi:10.1177/0146167202282006 0011000006286345 Schmitt, M. T., Branscombe, N. R., & Postmes, T. (2003). Women’s Wheaton, B., Muthén, B., Alwin, D. F., & Summers, G. (1977). Assessing emotional responses to the pervasiveness of gender discrimination. Eu- reliability and stability in panel models. Sociological Methodology, 8, ropean Journal of Social Psychology, 33, 297–312. doi:10.1002/ejsp 84–136. doi:10.2307/270754 .147 Whitley, B., Jr. (2009). Religiosity and attitudes toward lesbians and gay Schmitt, M. T., Spears, R., & Branscombe, N. R. (2003). Constructing a men: A meta-analysis. International Journal for the Psychology of minority group identity out of shared rejection: The case of international Religion, 19, 21–38. doi:10.1080/10508610802471104 students. European Journal of Social Psychology, 33, 1–12. doi: Ysseldyk, R., Haslam, S. A., Matheson, K., & Anisman, H. (2012). Love 10.1002/ejsp.131 thine enemy? Evidence that (ir)religious identification can promote Schumacker, R. E., & Lomax, R. G. (2004). A beginner’s guide to struc- outgroup tolerance under threat. Group Processes & Intergroup Rela-

This document is copyrighted by the American Psychological Association or one of its allied publishers. tural equation modeling (2nd ed.). Mahwah, NJ: Erlbaum Publishers. tions, 15, 105–117. doi:10.1177/1368430211410996

This article is intended solely for the personal use ofSellers, the individual user and is not to be disseminated broadly. R. M., & Shelton, J. (2003). The role of racial identity in perceived Ysseldyk, R., Matheson, K., & Anisman, H. (2010). Religiosity as identity: racial discrimination. Journal of Personality and Social Psychology, 84, Toward an understanding of religion from a social identity perspective. 1079–1092. doi:10.1037/0022-3514.84.5.1079 Personality and Social Psychology Review, 14, 60–71. Shen, M., Yelderman, L. A., Haggard, M. C., & Rowatt, W. C. (2013). Zuckerman, P. (2009). Atheism, secularity, and well-being: How the find- Disentangling the belief in God and cognitive rigidity/flexibility com- ings of social science counter negative stereotypes and assumptions. ponents of religiosity to predict racial and value-violating prejudice: A Sociology Compass, 3, 949–971. doi:10.1111/j.1751-9020.2009 Post-Critical Belief Scale analysis. Personality and Individual Differ- .00247.x ences, 54, 389–395. doi:10.1016/j.paid.2012.10.008 Sherkat, D. (2008). Beyond belief: Atheism, agnosticism, and theistic certainty in the United States. Sociological Spectrum, 28, 438–459. Received November 26, 2013 doi:10.1080/02732170802205932 Revision received August 1, 2014 Silberman, I. (2005). Religious violence, terrorism, and peace: A meaning Accepted August 28, 2014 Ⅲ