Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Evolved danger avoidance and implicit racial stereotypes effects of ambient darkness and beliefs about… Park, Justin H. 2002

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2002-0526.pdf [ 2.07MB ]
JSON: 831-1.0099673.json
JSON-LD: 831-1.0099673-ld.json
RDF/XML (Pretty): 831-1.0099673-rdf.xml
RDF/JSON: 831-1.0099673-rdf.json
Turtle: 831-1.0099673-turtle.txt
N-Triples: 831-1.0099673-rdf-ntriples.txt
Original Record: 831-1.0099673-source.json
Full Text

Full Text

E V O L V E D D A N G E R A V O I D A N C E A N D IMPLICIT R A C I A L STEREOTYPES: EFFECTS OF AMBIENT DARKNESS A N D BELIEFS A B O U T D A N G E R by JUSTIN H . P A R K B.Sc. The University of Alberta, 2000 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF M A S T E R OF ARTS in THE F A C U L T Y OF G R A D U A T E STUDIES (DEPARTMENT OF PSYCHOLOGY) We accept this thesis as conforming to the required standards THE UNIVERSITY OF BRITISH C O L U M B I A August 2002 © Justin H. Park, 2002 U B C Rare Books and Special Collections - Thesis Authorisation Form Page 1 of 1 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r a n a d v a n c e d d e g r e e a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l m a k e i t f r e e l y a v a i l a b l e f o r r e f e r e n c e a n d s t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s m a y b e g r a n t e d b y t h e h e a d o f m y d e p a r t m e n t o r b y h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t b e a l l o w e d w i t h o u t m y w r i t t e n p e r m i s s i o n . T h e U n i v e r s i t y o f B r i t i s h C o l u m b i a V a n c o u v e r , C a n a d a D e p a r t m e n t 7/16/2002 11 Abstract Much research suggests that perception of intergroup conflict, threat, and heightened danger may exacerbate outgroup derogation (including racial stereotyping). Based on evolutionary reasoning, ambient darkness and individual differences in belief in a dangerous world (BD W; Altemeyer, 1988) were expected to influence the automatic activation of African stereotypes connoting danger. The implicit association test (IAT; Greenwald, McGhee, & Schwartz, 1998) was used to measure the automatic activation of implicit stereotypes. A n interactive effect of B D W and darkness was found. Specifically, there was a positive correlation between participants' B D W and the extent to which participants associated African with the category "danger," but only under conditions of ambient darkness. B D W was unrelated to the extent to which participants associated African with the category "unpleasant." These results support the evolutionary psychological framework, which offers a coherent theoretical framework for understanding the nature of the relationship between danger-perception and outgroup derogation. Some limitations and implications of the study are discussed. i i i T A B L E OF CONTENTS Abstract i i Table of Contents i i i List of Tables v List of Figures vi Acknowledgements vii Introduction 1 Automatic Stereotyping 2 Perception of Danger and Derogatory Stereotyping 3 The Evolutionary Perspective 5 Intergroup Vigilance Theory 6 Effects of Belief in a Dangerous World (BDW) 9 Effects of Ambient Darkness 9 Interactive Effects 10 Overview of the Present Study 10 Method 12 Participants 12 Pretesting and Selection of Stimulus Items 13 IAT Task 14 Procedure 16 Results 17 Individual Difference Measures 17 IAT Effects 18 Effects of Ambient Darkness and B D W 19 iv Discussion 21 Conceptual Implications 22 Possible Limitations and Other Issues 22 Methodological Limitations 23 Conceptual Limitations 24 Culture and Gender 24 Further Comments on the Evolutionary Perspective 26 Cautionary Remarks 26 Domain Specificity 27 The Relationship between Stereotyping and Prejudice 27 Practical Implications 29 Future Directions and Concluding Remarks 29 References 31 Footnotes 36 Tables 37 Figures 41 List of Tables Table 1. Correlations among the individual difference measures, (p. 37) Table 2. Mean IAT effects of high and low B D W participants in dark and light conditions, (p. 38) Table 3. A N O V A results on effects of darkness and B D W on the unpleasant-pleasant and danger-safety IAT effects, (p. 39) Table 4. Results of regression analyses on effects of darkness and B D W on the unpleasant-pleasant and danger-safety IAT effects, (p.40) vi List of Figures Figure 1. Schematic illustration of the implicit association test (IAT): A n example of the unpleasant-pleasant IAT taken by an Asian participant, (p. 41) Figure 2. Unpleasant-pleasant IAT effects among high- and low-BDW participants in the Light and Dark conditions, (p. 42) Figure 3. Danger-safety IAT effects among high- and low-BDW participants in the Light and Dark conditions, (p. 43) Figure 4. Correlation between B D W and unpleasant-pleasant IAT effect, (p. 44) Figure 5. Correlation between B D W and danger-safety IAT effect, (p. 45) V l l Acknowledgements I would like to thank the members of my thesis committee, Jennifer Campbell, Steven Heine, and Mark Schaller, for their guidance and input in the completion of my thesis and my master's program. I would especially like to thank my supervisor, Mark Schaller, who continues to teach me the essentials of being an academic psychologist, researcher, and scientist. I also thank Simrat Sarahan for her assistance with data collection, Jason Faulkner for his intellectual companionship, Daniel Song for his helpful comments on my writing style, and my friends and family who provided the needed escape from school. Lastly, I would like to thank my parents, who always encouraged me to think independently, and without whom none of my achievements would be possible. 1 Research in intergroup cognition has identified the perception of danger as an important underlying factor of racial stereotyping. Danger may be perceived during times of anticipated or actual intergroup conflict, when one's self-image is threatened in some way, or when one feels vulnerable to physical harm, either dispositionally or as a result of situational factors. Indeed, there is some indication that such instances of perceiving danger or threat increase the incidence of derogatory racial stereotyping, usually accompanied by exacerbated prejudice against the target group (e.g., Altemeyer, 1988; Campbell, 1965; Schaller & Park, in press; Stephan, Ybarra, Martinez, Schwarzwald, & Tur-Kaspa, 1998). It seems logical that derogation of a specific racial outgroup increases as a function of the threat posed by that particular outgroup. Less apparent is how derogatory stereotyping of various racial groups is influenced by factors such as individual differences in wariness of dangers in general, situational cues that exacerbate the danger that is perceived, and cognitive factors that process information about groups and danger in systematic ways. The present paper attempts to provide a conceptual link between danger-perception and various instances of racial stereotyping. Described below is an experimental investigation that provides some support for the theoretical framework, and adds another important piece to the increasingly sophisticated body of knowledge on stereotyping. More specifically, the study examined the effects of ambient darkness and chronic wariness of danger on the automatic activation of the danger-relevant domain of African stereotypes. Before describing the study there are some issues that need clarifying and some conceptual background that must be highlighted. In the following sections are: (a) a description of the automatic nature of stereotyping, (b) a review of studies suggesting that there is indeed a relationship between perceiving danger and derogatory stereotyping, and (c) an outline of the evolutionary psychological framework from which the specific hypotheses for the present investigation were derived. 2 Automatic Stereotyping Much evidence suggests that stereotypes are often activated automatically, outside of the perceiver's awareness and control. Consequently, attempts have been made to understand the relationship between the automatic component and components that appear to be under the perceiver's control. Devine (1989) argues that everyone automatically activates stereotypes to more or less the same extent in the mere presence of a member of the relevant group. She argues further that some individuals (those identified as high-prejudiced) have personal belief structures that happen to overlap with the automatic stereotypes, and thus, overtly express views consistent with their automatically activated prejudice and stereotypes. Others (those identified as low-prejudiced), however, hold personal beliefs that are contradictory to the stereotypes, and thus, exercise control over their overt expression. Essentially, the automatic and controlled components are seen as independent (although the automatic component is seen as a necessary sequential step). While the details of her model have been contested—in particular, the idea that unintentional activation of stereotypes is uniform across individuals— the proposition that stereotyping has an automatic component is well supported (Banaji & Hardin, 1996; Blair & Banaji, 1996; Lepore & Brown, 1997; Wittenbrink, Judd, & Park, 1997). Contrary to Devine's (1989) original proposal, others have found evidence that stereotypes are not uniformly activated across all individuals under all circumstances. The basis of this notion is that automatic activation of stereotypes involves cognitive associations whose strength is conditional (Blair, Ma, & Lenton, 2001). Indeed there is evidence that automatic activation of stereotypes is contingent upon various factors. For instance, Blair and her colleagues (Blair et al.) found that participants who were instructed to imagine a strong woman (which is counterstereotypic) were substantially less likely to automatically activate the "weak female" stereotype. 3 Other studies provide evidence that automatic stereotyping depends on the perceiver's intention (Blair & Banaji, 1996), whether category (e.g., Black) or stereotype (e.g., hostile) is used as prime (Lepore & Brown, 1997), and exposure to stereotype-consistent and inconsistent exemplars (Dasgupta & Greenwald, 2001). Also, some investigators have interpreted the variability in automatic stereotyping as an indicator of individual differences in (implicit) attitudes toward racial groups (Fazio, Jackson, Dunton, & Williams, 1995; Greenwald, McGhee, & Schwartz, 1998). Thus, under the assumption that automatic activation of stereotypes is malleable and contingent, the present study investigated the extent to which the automatic activation of racial stereotypes is influenced by specific variables. The recognition that automatic stereotyping is malleable has led to some creative attempts to measure the magnitude of automatic stereotyping, many of them utilizing subliminal priming and/or reaction time methods. The present investigation used one such method called the implicit association test (IAT; Greenwald et al., 1998), which is described in detail below. In addition to the conceptual clarity and rigor, examining the automatic component of stereotyping confers the added advantage that participants' responses remain largely immune to socially desirable responding and demand characteristics. Perception of Danger and Derogatory Stereotyping With respect to the impact that the perception of danger has on derogatory stereotyping, the most obvious of such instances occurs in times of anticipated or actual conflict between two groups—particularly between one's ingroup and an outgroup. In his realistic conflict theory, Campbell (1965) argues that outgroup derogation occurs partly as a result of perceiving actual threat posed by specific outgroups. This idea has received empirical support in a study of people's judgments of immigrant groups (Stephan et al., 1998). Studies with ad-hoc groups consistently show that anticipated intergroup conflict leads to greater amplification of intergroup differences on various evaluative characteristics (see review by Brewer, 1979). The 4 psychological consequences of actual conflict were nicely demonstrated by Sherif and colleagues (Sherif, Harvey, White, Hood, & Sherif, 1961) in a study involving groups of boys divided into two groups. After a few days of intergroup competition, the boys quickly displayed signs of outgroup derogation, referring to the other group as being "dirty" and "rough." Threat to one's self-image has also been shown to influence stereotyping. Fein and Spencer (1997) demonstrated that when one's self-image is threatened, individuals are more likely to display stereotyping of and prejudice against stigmatized outgroups. This effect occurred even when participants were cognitively busy (Spencer, Fein, Wolfe, Fong, & Dunn, 1998), suggesting that self-image threat influences stereotyping even at the automatic level. However, because this process involves perceiving somewhat of a "symbolic" danger, the relationship between this process and those that involve actual physical danger is not entirely clear. Perception of an elevated level of danger in the environment (not associated with any specific outgroup) also seems to influence derogation of various outgroups. Evidence suggests that an elevated wariness of generalized dangers in the environment may contribute to derogation of specific outgroups that have played no part in creating that original wariness (Altemeyer, 1988). Although empirical evidence is scarce, we are well aware of people's tendency to lash out against various outgroups during times of international strife. Often, the target of such reactivity is a group that is objectively unrelated to the problems—a "scapegoat." More evidence is found in studies of individual differences: Altemeyer found that people who are chronically wary of dangers in the world are more likely to display racial prejudice. Additionally, fearfulness is a key construct of the authoritarian personality (Adorno, Frenkel-Brunswik, Levinson, & Sanford, 1950), which is also related to prejudicial tendencies. Within this personality and social-cognitive framework, the present study investigated the combined effects of a situational variable and an individual difference variable on the 5 automatic activation of danger-relevant African stereotypes. The rationale for examining these particular factors and domain is derived, at least in part, from evolutionary psychological reasoning. The Evolutionary Perspective The social-cognitive approach is amenable to the evolutionary psychological perspective in that they both view the mind as being modular to some extent; that is, they both acknowledge the presence of multiple domain-specific psychological mechanisms (Tooby, & Cosmides, 1989). A major contribution of the evolutionary perspective is the identification of mechanisms that may have served an adaptive function in the past and that may still persist today (Buss, 1995). The logical steps of this perspective are: (a) based on some assumptions regarding the ancestral environment, identify specific adaptive problems that such an environment may have posed on ancestral humans; (b) propose psychological mechanisms that may have evolved in response to these specific problems; and (c) deduce some key design features of the evolved mechanism, such as identifying the circumstances under which the mechanism was (and is) most likely to be activated. A key utility of this perspective is its ability to generate hypotheses regarding how an evolved mechanism may function in the contemporary environment, hypotheses that can then be tested using standard experimental psychological methods. In the present paper, it is argued that some instances of racial stereotyping may have their roots in a psychological mechanism that, in ancestral environments, served a specific adaptive function. A n important implication is that this particular mechanism may be logically independent from those that underlie stereotyping other potential outgroups, such as women or the elderly. Likewise, stereotyping pertaining to one domain (e.g., "they are dangerous") may be logically independent from that pertaining to other domains (e.g., "they are lazy, ignorant, etc."). Consequently, while stereotypes pertaining to different types of outgroups and domains 6 may share some common features, they may be logically independent in some important ways, and thus, influenced by different sorts of personality and contextual factors. This is the rationale for limiting the scope of this research to racial groups and to one particular domain of outgroup derogation. Intergroup vigilance theory. Intergroup vigilance theory (Schaller, in press, 2001; Schaller & Park, in press) suggests a possible mechanism of outgroup derogation that may pertain to racial outgroups and stereotypes that pertain to danger. The reasoning behind this theory is as follows: Given that ancestral humans lived in relatively small groups, and that unexpected contact between these groups was often dangerous, any cognitive process that facilitated the vigilant avoidance of such encounters would have been adaptive. And one process that may have assisted in the avoidance of these tribal outgroup members would have been the tendency to exaggerate the level of danger posed by these outgroup members. (For the remainder of the paper, the mechanism hypothesized to underlie this tendency wil l be called the danger-avoidance mechanism.) The implication is that in the presence of appropriate triggers (see below), the danger-avoidance mechanism may be activated, leading to exaggerated perceptions of certain outgroup members as being hostile or untrustworthy. Adaptive mechanisms are not activated indiscriminately, however (Schaller, in press). Rather, they are typically activated in situations in which they confer the most adaptive benefit (i.e., benefits outweighed the costs). And according to Schaller, activating this danger-avoidance mechanism would have been most beneficial (a) in situations in which the likelihood of unexpected intergroup interaction was especially high, and (b) in situations in which an intergroup interaction was especially likely to be violent. How would ancestral humans have recognized these situations? To the extent that various heuristic cues reliably accompanied these situations, one plausible way would have been to evolve to respond to these cues. There were likely heuristic cues that signaled an increased chance of contact with outgroup members 7 or an increased likelihood of an interaction being dangerous—the very situations in which perceiving outgroup members as being especially dangerous would have been most beneficial. And reacting appropriately to such cues—perceiving outgroup members as being especially dangerous, and thus, avoiding them—would have been adaptive. If this particular danger-avoidance mechanism is distinct and independent, the heuristic cues should have lead ancestral humans to perceive tribal outgroup members as especially dangerous, but they should not have directly influenced perceptions of outgroup members on derogatory trait dimensions that are unrelated to danger (e.g., intelligence, cleanliness, etc.). Though the world has changed drastically since the ancestral times in which our psychological mechanisms evolved, because there has been insufficient time for any major change in genotype, it is reasonable to suppose that some of these mechanisms still function in more or less the same way regardless of whether they are conferring adaptive benefits in the contemporary environment (Tooby & Cosmides, 1989). Thus, once we deduce the ways in which a potential psychological mechanism may have functioned in the past, we can generate hypotheses about how the mechanism might respond in the present. Having identified the relevant outgroups and the heuristic cues that might activate and influence the mechanism, it becomes possible to experimentally investigate whether these proposed mechanisms in fact exist, and whether they respond in the proposed manner. If the danger-avoidance mechanism proposed by intergroup vigilance theory in fact exists, it would apply specifically to contemporary intergroup contexts that are analogous to the "tribal" context (Schaller, 2001). In other words, social outgroups that resemble the tribal outgroups along some relevant dimensions should activate the mechanism. Some contemporary social groups appear to fit the template more closely than others. As suggested by Schaller, "ingroup/outgroup distinctions based on race, ethnicity, and geopolitical boundaries" (p. 16) 8 appear to be analogous to the tribal context. In contrast, distinctions based on gender, age, and social class do not fit the template as closely. With respect to heuristic cues, being chronically wary of dangers may potentially be a heuristic cue signaling that "something is wrong," resulting in a greater likelihood of the danger-avoidance mechanism being activated, prompting the individual to perceive an outgroup as being especially dangerous. Indeed, as mentioned above, the tendency to believe that the world is dangerous has been shown to be related to outgroup derogation (Altemeyer, 1988). Additionally, various events in the environment or the presence of temporary situational cues may signal an elevated likelihood of contact with outgroup members or a higher likelihood of an interaction being dangerous. One such event or cue might be the presence or anticipation of some sort of a realistic threat or conflict (Brewer, 1979; Campbell, 1965). Other potential cues are ambient darkness, outgroup size, and outgroup proximity (Schaller, 2001). This theoretical framework makes sense of some of the factors that are known to contribute to outgroup derogation and stereotyping. Furthermore, outgroup derogation processes that are logically independent from that outlined by the theory are explicitly omitted. For instance, processes underlying outgroup derogation resulting from self-image threat, derogatory stereotyping on trait dimensions unrelated to danger, and derogation of outgroups far removed from the tribal template are seen to be logically independent from—and outside the scope of—the process outlined by intergroup vigilance theory. The purpose of the present study was to investigate the combined effects of two factors on the automatic activation of stereotypes connoting danger. The first is a personality variable already mentioned above—being chronically wary of dangers in the world, or belief in a dangerous world (BDW; Altemeyer, 1988). Although Altemeyer's results revealed a relationship between B D W and prejudice, they did not show that B D W is related specifically to 9 danger-relevant stereotypes. The second is a contextual cue derived from intergroup vigilance theory: ambient darkness (Schaller, 2001). Effects of belief in a dangerous world (BDW). There is some evidence that individuals scoring higher on the B D W measure display greater activation of derogatory stereotypes that connote danger on explicit, self-report measures. In one study (Schaller, Kelly, & Dhanoa, 2001), Canadian participants scoring higher on B D W expressed more anti-immigration attitudes specific to an African group, as well as anti-immigration attitudes in general. Interestingly, these attitudes seemed to be related to beliefs that the outgroup possesses danger-relevant traits, but not other derogatory traits. That is, individuals with higher B D W scores rated the outgroup as being more hostile and less trustworthy, but B D W did not predict ratings on other derogatory traits that are unrelated to danger. Effects of ambient darkness. Ambient darkness is known to have various perceptual effects. Interestingly, darkness has been shown to increase the magnitude of the acoustic startle reflex in humans (Grillon, Pellowski, Merikangas, & Davis, 1997), perhaps related to the danger-avoidance process. One investigation examining the effects of ambient darkness suggests that people may be more aggressive in the dark (Page & Moss, 1976). Within the framework of intergroup vigilance theory, ambient darkness may be an important heuristic cue that, through visual impairment, signals an elevated likelihood of unexpected contact with unknown people and a higher likelihood of the encounter being dangerous (i.e., ambient darkness is a cue that connotes danger). If this is the case, ambient darkness should increase the tendency for people to perceive racial outgroup members as being especially dangerous. This is seen in a few studies. In one study (Schaller & Park, in press), Canadian high school students rated people from Iraq and people from Canada on four trait dimensions: hostile, ignorant, trustworthy, and open-minded (two equally negatively-valenced and two equally positively-valenced). Two of 10 the traits—hostile and trustworthy—are danger-relevant, while the other two are not. Approximately half of the participants completed the ratings in an illuminated room, and the rest completed the ratings in a room in which the lights had been turned off to create near-total darkness. The results showed that the ingroup (Canadians) was favored along all four trait dimensions. Interestingly, being in the dark seemed to increase the differential ingroup-outgroup ratings on the danger-relevant traits, while darkness had no impact on the differential ingroup-outgroup ratings on the danger-irrelevant traits. In another study (Schaller et al., 2001), participants listened to information about a potential immigrant group on audiotape in either a well-lit room or in near-total darkness. Results revealed that attitudes toward immigration were more negative in the dark. Interactive effects. It seems reasonable that these cues might have interactive effects such that the presence of both cues further exacerbates derogatory stereotyping. A study by Schaller et al. (2001) revealed interactive effects of B D W and ambient darkness on Canadian students' ratings of Iraqis on danger-relevant traits. Specifically, darkness increased the tendency of high-BDW participants (but not low-BDW participants) to perceive Iraqis as more dangerous. The studies described above provide some support for the evolutionarily derived hypotheses. At the very least, they strongly suggest that ambient darkness and B D W may indeed have specific effects on the way people view outgroup members. Common to all of these studies is that participants exercised full control over their responses. As discussed earlier, much of stereotyping occurs outside of the perceiver's awareness and control. Hence, the results of these studies may pertain only to the controlled component of stereotyping. Overview of the Present Study The purpose of this study was to replicate the previous findings at the automatic level of stereotyping. The method that was used to measure the automatic activation of stereotypes and 11 attitudes in the present study is the widely used implicit association test (IAT; Greenwald et al., 1998). The IAT is a computer-based reaction time method that measures differential association of two social categories (e.g., African-European) with attributes along an evaluative dimension (e.g., unpleasant-pleasant). The IAT has been used successfully to measure the automatic activation of various stereotypes and attitudes (e.g., Ashburn-Nardo, Voils, & Monteith, 2001; Greenwald et al., 1998; Rudman, Greenwald, Mellott, & Schwartz, 1999). Importantly for the purposes of the present study, responses on the IAT have been shown to be sensitive not only to chronic differences in associative strength, but to temporary shifts in associative strength as well (Blair et al., 2001; Dasgupta & Greenwald, 2001; Lowery, Hardin, & Sinclair, 2001; Wittenbrink, Judd, & Park, 2001). Thus, it is a useful tool for assessing the effects of a situational variable and/or an individual difference variable on the implicit association between a social group and a trait category, which serves as a measure of the strength of automatically activated stereotypes. Additionally, this methodology is immune to participants' attempts to control their responses. In the present study, the racial ingroup for participants of East Asian heritage was "Asian," and for the rest of the participants (predominantly of European heritage), the ingroup was "European." Africans were chosen as the racial outgroup in this study for a couple of reasons. First, because virtually none of the participants at the University of British Columbia are of African heritage, they provide a salient racial outgroup for most of the participants. Second, African Americans are a group whose stereotypes contain traits connoting danger, along with other derogatory traits that do not connote danger (Devine, 1989), and thus, likely to reveal differential activation of danger-relevant stereotypes. Also, all target individuals were males. As Schaller (2001) notes, this particular danger-avoidance tendency may plausibly be present most robustly in male perceivers with male 12 targets. Thus, this investigation focused only on male targets, although both males and females were included in the perceiver role. Two separate IAT tasks were created—one measuring the association between African and unpleasant (the unpleasant-pleasant IAT), and a second measuring the association between African and danger (the danger-safety IAT). Participants were randomly assigned to one of two Darkness conditions: the Dark condition and the Light condition. Participants also completed the B D W measure. The B D W questionnaire is composed of 12 items, and participants are asked to indicate—on a 7-point scale—the extent to which they agree with statements such as "There are many dangerous people in the society who will attack someone out of pure meanness, for no reason at all." It was predicted that participants in the Dark condition and/or participants scoring higher on B D W would display a stronger danger-safety IAT effect. The unpleasant-pleasant IAT effect was not expected to be as substantially influenced by Darkness or BDW. A number of other individual difference measures that are known to be related to prejudice and stereotyping in various contexts were also administered. There was no theoretical basis on which to hypothesize differential relations between these personality variables and the two IAT tasks. In fact, an absence of such relations was expected based on the hypothesized domain-specific relationship between B D W and the activation of danger-relevant stereotypes. Method Participants Fifty-three undergraduate students participated in exchange for course credit in first-and second-year psychology courses. Forty-two participants were women and 11 were men; 26 were of East Asian heritage, 18 were of European heritage, 2 were of both Asian and European heritage, and 7 were of various other ethnic backgrounds (none were of African heritage). Data from two participants were omitted during analysis; one did not complete the relevant 13 questionnaires, and one had a high error rate on a relevant block of the IAT (less than 80 percent correct responding). Thus, data from the remaining 51 participants were analyzed. Pretesting and Selection of Stimulus Items Forty additional participants in a preliminary study were given a three-part questionnaire consisting of a list of 50 words that have either positive or negative connotations (including words used in prior IAT studies by Greenwald et al., 1998). Participants rated on a 7-point scale the extent to which the words are associated with feelings of either unpleasantness or pleasantness (1 = unpleasant; 4 = neutral; 7 = pleasant). Second, they rated the extent to which the words are associated with danger and/or threat. Third, they rated the extent to which the words are associated with safety and/or security. The last two ratings were made on a 7-point scale with the endpoints of 1 and 7 labeled "not associated" and "highly associated." The danger ratings were subtracted from the safety ratings to produce an index of danger-relevance, such that words that were perceived as connoting greater danger had more highly negative values, words that were perceived as connoting greater safety had more highly positive values, and words that were perceived as having little relevance for danger or safety had values closer to 0. (E.g., "murder" had a value of-5.48, "secure" had a value of+4.72, and "lazy" had a value of-0.43.) On the basis of these pretest ratings, stimulus words for two different IAT tasks were selected. One IAT was designed to assess differential associations of racial groups with unpleasant or pleasant; the other was designed to assess associations with danger or safety. For the unpleasant-pleasant IAT, 5 unpleasant (dirty, failure, lazy, poverty, stink) and 5 pleasant (intelligent, laughter, pleasure, rainbow, smart) words that were not rated as highly associated with danger or safety were selected. (Mean pleasantness ratings were 2.30 and 5.76, respectively; Af s on the danger-relevance index were -1.52 and 2.26, respectively.) For the danger-safety IAT, 5 unpleasant (aggressive, hurt, injury, pain, terrible) and 5 pleasant (gentle, 14 harmless, peace, secure, trust) words that were rated as highly associated with danger or safety were selected. (Mean pleasantness ratings were 2.33 and 5.63, respectively; Afs on the danger-relevance index were -3.52 and 3.87, respectively.) Thus, the two IAT tasks equally measured associations between outgroup-ingroup and evaluatively negative and positive words. They differed in that the danger-safety IAT was designed to measure more specific associations between outgroup-ingroup and danger-safety. As already mentioned, all target individuals were males. Their black-and-white pictures were obtained from high school yearbooks. For each picture, the mouth area was removed in order to eliminate much of the facial expressions. Thirty pictures were prepared to total, 10 for each of the racial groups (Asian, European, and African). IAT Tasks Each IAT task presented participants with a series of forced-choice categorization trials, in which they responded by pressing either the " E " or the "I" key on a keyboard with their left and right index fingers, respectively. On each trial, a target stimulus (e.g., a word) was presented at the center of the computer screen, with the two response option reminders (e.g., "unpleasant" and "pleasant") indicated on the upper left and upper right sides of the screen. Participants were instructed to respond as quickly as possible by pressing the response key corresponding to the category associated with the target stimulus. On the unpleasant-pleasant IAT (see Figure 1), participants were presented with two different kinds of categorization tasks. One type was a word categorization task in which participants judged whether words were unpleasant or pleasant. A second type was a racial outgroup-ingroup picture categorization task. Participants of East Asian heritage judged whether faces were African or Asian, and all other participants (those of European, mixed, and other backgrounds) judged whether faces were African or European. 15 These trials were presented across five distinct blocks, each of which was prefaced with a different set of instructions. The first block presented 20 word categorization trials, and the second block presented 20 face categorization trials. The third block presented 40 trials, randomly mixing word categorization and face categorization trials, in which unpleasant was paired with the same response key as the racial outgroup (African), and pleasant was paired with the same key as the ingroup (Asian or European). Following another block of just 20 face categorization trials (in which the response keys associated with ingroup and outgroup were reversed), the fifth block presented 40 trials mixing word and face categorization trials. In contrast to the third block, unpleasant was paired with the same response key as the ingroup (Asian or European), and pleasant was paired with the same key as the outgroup (African). Previous research (e.g., Greenwald et al., 1998) has shown that the difference in average response times to trials presented in Block 3 and Block 5 serves as an indicator of the extent to which the outgroup (African) is differentially associated with the evaluative attribute unpleasant and the ingroup is associated with the evaluative attribute pleasant. That is, shorter response times at Block 3 relative to response times at Block 5 serves as a measure of implicit cognitive association between African and unpleasant. The danger-safety IAT followed the same format, except that the stimulus words were those rated to be particularly relevant to either danger or safety, and the response categories that participants used on the word categorization trials were labeled "danger" and "safety" (participants were instructed to judge whether words were danger words or safety words). The difference in average response times on Block 3 and Block 5 serves as an indicator of the extent to which African is differentially associated with danger. The two IAT tasks were presented sequentially, with no interruption from the experimenter, and the order of the two tasks was counterbalanced across subjects. 16 IAT studies that focus solely on the absolute magnitude of IAT effects (e.g., Greenwald, et al., 1998; Rudman et al., 1999) need to counterbalance the order in which the outgroup is first associated with unpleasantness or pleasantness (Blocks 3 and 5). In the present study, the absolute magnitude of the IAT effects was of no interest; rather, the purpose was to examine the effects of additional independent variables on the IAT effects, which served as the dependent variable. Thus, no such counterbalancing was done.1 IAT tasks were presented on an I B M compatible computer through a program run on Inquisit software (Version 1.29). Participants viewed the display from a distance of approximately 60 cm. Procedure At the outset of the experimental session, participants were seated in a small room in front of a computer, and were presented with a practice IAT that familiarized them with the IAT procedure. (The categorization trials in this practice IAT involved pictures of insects and flowers, and unpleasant and pleasant words; these words were different from those used in the unpleasant-pleasant IAT described above.) Participants were then taken into a different room to complete a demographic questionnaire and 2 other questionnaires: one measuring personal need for structure (PNS; Neuberg & Newsom, 1993) and another measuring need for cognition (NFC; Cacioppo & Petty, 1982). Upon completion of the questionnaires, participants were informed that they would be doing another IAT task shortly. Prior to proceeding with this task, the darkness manipulation was introduced. The procedural context of this manipulation was somewhat different for different participants. Thirty participants were told that the researchers were interested in studying perceptions of different ethnic groups, in particular African Americans, and that they should sit for 30 seconds thinking about African Americans. The experimenter then left 17 participants alone in the room with the door closed. The windowless room in which they were left alone was either well lit by overhead fluorescent lights (Light condition), or the lights were turned off so that the room was completely dark (Dark condition). The other 23 participants were not asked to think specifically of any ethnic group. They were simply told that before they proceeded with the next task, they should clear their minds for 30 seconds. The experimenter then left participants alone in the windowless room with the door closed. Again, the room was either well lit by overhead fluorescent lights (Light condition), or the lights were turned off (Dark condition). The essential difference between these two versions of the darkness manipulation was whether participants were asked specifically to think about African Americans or not.2 The results were unaffected by this distinction, and so participants' responses were pooled into two overall Light and Dark conditions. Following the 30 seconds in which participants were left alone in either the light or the dark, the experimenter returned and started participants on the IAT tasks that assessed stereotypic perceptions of Africans. Each participant completed the IAT alone with the door closed, and with the overhead lights either turned on or off (depending on whether they were in the Light or Dark condition). Upon completion of the IAT, participants completed the questionnaires measuring motivation to control prejudiced reactions (MCPR; Dunton & Fazio, 1997), belief in a dangerous world (BDW; Altemeyer, 1988), and perceived vulnerability to disease (PVD; Schaller et al., 2002). Participants were then debriefed, thanked, and dismissed. Results Individual Difference Measures The internal consistencies (Cronbach's alpha) of the individual difference measures and the correlations among them are shown in Table 1. Not surprisingly, the results revealed moderate correlations between B D W and PVD, both of which involve some sort of physical 18 threat to the self, and between PNS and MCPR, which involve a desire for structure and control. Because the MCPR, BDW, and P V D measures were administered following the IAT tasks (in which participants had been assigned to one of two Darkness conditions), the means of these measures were compared across the two conditions to examine possible carry-over effects of the Darkness manipulation. Results revealed that none of these means differed across the two conditions (allp's > .40). IATEffects The two IAT tasks can be used to produce two different indicators of the extent to which Africans (compared to the racial ingroup) are differentially associated with evaluatively negative characteristics. The unpleasant-pleasant IAT yields an index of the extent to which Africans are associated with generally unpleasant characteristics. The danger-safety IAT yields an index of the extent to which Africans are associated with the more specific set of evaluatively negative characteristics connoting danger. To compute these two indices, the following steps were taken on the data generated by each of the two IAT tasks: (a) Mean reaction times were computed within Blocks 3 and 5. In order to eliminate any unusually fast or slow reaction times that can occur when participants reoriented to the task at the beginning of each block of trials, responses to the first two trials at each block were omitted when computing these means. (b) The mean reaction time on Block 3 (for which the African response key was identical to the unpleasant or danger response key) was subtracted from the mean reaction time on Block 5 (for which the African response key was identical to the pleasant or safety response key). If the resulting value is positive, it indicates that the social category African is differentially associated with unpleasant (for the unpleasant-pleasant IAT) or with danger (for 19 the danger-safety IAT). The magnitude of this value indicates the magnitude of this stereotypic association. Consistent with previous IAT studies assessing racial stereotypes, results indicated a general tendency to differentially associate Africans with unpleasant characteristics. Across all participants, the mean differential response time on the unpleasant-pleasant IAT was 121.61, a value that is clearly different from zero, /(50) = 3.80, p < .001. On the danger-safety IAT, the mean differential response time was even more strongly positive, M= 141.51,?(50) = 5.35,/?< .001, indicating that participants tended to differentially associate Africans with danger. The magnitude of the two IAT effects did not differ, t(50) = .62, p = .54. As mentioned before, the method did not include counterbalancing of the order in which the outgroup was paired with either the evaluatively negative or positive trait categories. Thus, these particular IAT effects by themselves are not diagnostic of the absolute magnitude of the cognitive associations that participants made between the outgroups and the evaluatively negative trait categories. In this study, the question was not whether participants would reveal IAT effects, but whether these effects would be influenced by ambient darkness and/or individual levels of BDW. There were also differences in the magnitude of the IAT effects between participants of East Asian and European background. Specifically, East Asians displayed stronger IAT effects for both the unpleasant-pleasant IAT, t(4\) = 2.67, p = .01 (Ms were 220.26 and 41.63 for East Asians and Europeans, respectively), and for the danger-safety IAT, t(4l) - 2.37,p = .02 (Ms were 187.63 and 58.08). However, the effects of Darkness and B D W did not appear to differ between these two groups (the N's were too small at this point to draw any conclusive inferences), thus, data from all participants were pooled together. Effects of Ambient Darkness and BDW There are several ways to examine whether the IAT effects were influenced by B D W and ambient darkness. For a rudimentary analysis, the B D W scores were separated into two 20 groups via a median split (high and low BDW). In this 2 (BDW: high or low) X 2 (condition: dark or light) factorial design, means of the IAT effects are shown in Table 2. Two separate 2 X 2 A N O V A ' s were conducted, one in which the unpleasant-pleasant IAT effect was the dependent variable, and another in which the danger-safety IAT effect was the dependent variable. The results of these A N O V A ' s are displayed in Table 3. On the unpleasant-pleasant IAT, neither Darkness nor B D W had main effects, nor was there any indication of an interaction effect (see Figure 2). On the danger-safety IAT, Darkness and B D W did not have main effects, but there was a significant interaction effect (see Figure 3). Analyses of simple main effects for the danger-safety IAT effect revealed that among low-BDW participants, the Darkness manipulation had a significant effect, F ( l , 47) = 7.60, p < .01, such that those in the Light condition (M-226.45) showed substantially stronger IAT effects than those in the Dark condition ( M = 20.39); and among high-BDW participants, IAT effects in the Light condition (M= 137.69) and in the Dark condition (M= 179.44) did not differ, F(l,47) = .38, p > .50. This curious finding of low-BDW participants displaying greater danger-safety IAT effects in the Light (which seems to have driven the interaction effect) was unexpected. Rather, the prediction was that the Dark condition would increase the danger-safety IAT effect among high-BDW participants. Alternatively, within the Dark condition, B D W had a significant effect, F(l,47) = 5.26,p < .05, such that high-BDW participants (Af = 179.44) showed stronger IAT effects than low-BDW participants (M= 20.39); and within the Light condition, B D W did not have an effect, F(l,47) = 1.46,p> .10. Because B D W is a continuous variable, however, separating it into high and low groups may be an insensitive method that does not capture much of its influence on the IAT effects. Thus, two separate regression analyses were conducted on the two different IAT effects. For these analyses, the B D W and Darkness variables were first transformed into z-scores. An additional predictor variable corresponding to the B D W X Darkness interaction effect was 21 computed by taking the multiplicative products of the standardized B D W and Darkness variables. These three predictors were entered into a pair of regression analyses. Results of both regression analyses are presented in Table 4. The results revealed no significant effects for any of the predictors. However, the Darkness X B D W interaction predictor approached significance on the danger-safety IAT effect, B = -.26, p = .07, suggesting an interaction effect of Darkness and B D W on the implicit association of African with danger. Correlations between B D W and the two IAT effects were examined separately in the Light and Dark conditions. B D W did not correlate significantly with the unpleasant-pleasant IAT effect in either the Light (r = -.075, p = .73) or the Dark (r = .23, p = .25) condition (Figure 4). With the danger-safety IAT, B D W did not correlate significantly with the IAT effect in the Light condition (r = -.15, p = .50), but had a significant positive correlation with the IAT effect in the Dark condition (r = .39, p = .04; Figure 5). Chronic beliefs about danger did not predict the strength of the association between Africans and generalized unpleasantness in either the dark or the light. Beliefs about danger did predict the extent to which Africans were stereotypically associated with characteristics connoting danger, but only under conditions of ambient darkness. Also, none of the other individual difference measures correlated significantly with either of the IAT effects in either the Light or the Dark conditions. Discussion In contrast to some of the earlier research that focused on the "controlled" component of prejudice and stereotyping, this study failed to find any significant main effects of ambient darkness or BDW. Interestingly, a main effect that approached significance on both the A N O V A and regression analysis on the danger-safety IAT effect indicated that participants in the Light condition displayed stronger IAT effects. Also, the results of simple main effects suggest that most of the Darkness effect is among people low in BDW. That is, for high-BDW 22 individuals, the Darkness condition had virtually no effect on the danger-safety IAT effect, while for low-BDW individuals, those in the Light condition showed substantially stronger danger-safety IAT effects. This remains an uninterpretable, anomalous finding. The results of regression analyses suggest that B D W predicts automatic activation of danger-connoting stereotypes under conditions of ambient darkness, but not in the light. Also importantly, B D W and ambient darkness appear to have much less impact on the activation of African stereotypes that are not danger-relevant. Conceptual Implications Some implications follow from this study. First, the results suggest that the effects of B D W and ambient darkness on outgroup derogation may extend to the automatic component of stereotyping. Second, the results add support for domain-specificity (further discussed below); no other individual difference measure had any effects on the IAT effects. Perceived vulnerability to disease, which was moderately correlated with B D W (r = .31, p = .03), had no relation to either of the IAT effects. This suggests that the danger-avoidance mechanism is logically independent from—and thus, unrelated to—concerns about contagious disease. Third, the interactive effect of B D W and ambient darkness suggests that some dispositional variables may require an additional external variable to "trigger" their effects (Schaller & Park, in press). Possible Limitations and Other Issues While the results are mostly consistent with the hypotheses, there are several reasons— anomalous findings aside—to be cautious about what conclusions are drawn. Two broad domains of limitations will be addressed. First, there is much debate and controversy surrounding what exactly the IAT taps into. Critics of the IAT point out some reasons to remain wary of the findings. In addition, even if we assume that the IAT does measure implicit cognitive associations, there is further controversy regarding the relation between the implicit and explicit levels of outgroup derogation. Second, the present study examined only the case in 23 which Africans comprised the racial outgroup, which may impose limitations on the generalizability of the findings. Methodological limitations. Brendl, Markman, and Messner (2001) identify two major concerns regarding the IAT. First, there is evidence that participants recognize certain trials (e.g., outgroup associated with pleasant) as being more difficult than others (e.g., outgroup associated with unpleasant), and thus, shift their response criterion when doing the IAT. This has serious implications for the validity of the "IAT effect." Second, it is difficult to interpret the IAT data because, in principle, an IAT effect could emerge from ingroup favoritism alone, outgroup derogation alone, or a combination of both, and the IAT methodology is unable to dissociate the two. With respect to the present study, these concerns challenge the inference that the IAT effect represents an association that participants make between the racial outgroup (African) and the evaluatively negative trait categories, even though the hypotheses pertain mainly to outgroup derogation. That participants may be associating the ingroups with the evaluative positive categories cannot be ruled out. To obtain unequivocal evidence for the direction of the bias, methodologies that separate the processes could be used (such as the lexical decision task). Another major issue raised by critics of the IAT is how it relates to explicit measures of attitudes about outgroups, and whether it predicts behavior. Greenwald et al. (1998) report widely ranging correlations (from r = -.04 to r = .64) between IAT effects and explicit measures of prejudice. More recently, Karpinski and Hilton (2001) found that there is virtually no relation between the IAT and explicit measures of attitude, and found no evidence that the IAT predicts behavior. In addition, they report evidence supporting an "environment association model of IAT effects." That is, the IAT may be simply measuring associations that an individual is exposed to, rather than attitudes that the individual endorses, implicitly or otherwise. 24 Another potential limitation of this study is that the words that were used in the IAT were not all trait words. Words like "peace, laughter, failure, rainbow," while being evaluatively positive or negative, are not necessarily trait words that describe people. How this may be a problem is uncertain, but using trait words may help reduce ambiguity in interpreting the results. Also, the arguments of this paper would be stronger i f the association between African and danger was compared to association between African and another specific stereotypic trait (rather than general unpleasantness). Conceptual limitations. The present study explored essentially one direction of outgroup derogation; participants of Asian and European heritage completed the studies in which Africans comprised the outgroup. Based on the logic of intergroup vigilance theory, comparable results should be obtained with other combinations of ingroups and outgroups (e.g., African ingroup versus European outgroup, European ingroup versus Asian outgroup, etc.). However, numerous factors underlie the formation and maintenance of stereotypes, and stereotypes that connote danger appear to be fairly prominent among the African American stereotypes (Devine, 1989). The implication is that the effects of ambient darkness and beliefs about danger on activation of danger-relevant stereotypes may be weaker or non-existent with a different ingroup-outgroup pairing. In a way, this may be seen as a theoretical limitation. However, i f we acknowledge that stereotypes are multiply determined and that a number of causal factors underlie prejudice and stereotyping processes, then the fact that similar results would not be obtained with other ingroup-outgroup pairings is not necessarily a limitation. In fact, the results from this study may potentially contribute to a more complete understanding of derogation of Africans in particular. Culture and gender. The number of participants and the design of the present study did not allow a detailed exploration of possible effects of culture and gender on the IAT effects. 25 However, there are theoretical reasons to suppose possible culture and gender effects on the processes described in this paper. One obvious cultural factor is which group serves as the racial outgroup. Different sets of stereotypes are attached to different racial and ethnic groups, and as mentioned earlier, Africans have consistently been associated with traits connoting danger. In contrast, Asians in contemporary North America are viewed as the "model immigrants" (Fiske, 1998). We must also consider which group serves as the ingroup for the outgroup in question. Though there is a lack of research on this topic, we cannot suppose that Africans are perceived in the same manner by people of European and Asian heritages, among others. Cultural differences may run deeper. The recent burst of work in cultural psychology has provided ample evidence that many of the supposedly universal psychological mechanisms are, in fact, not so universal. Relevant to intergroup cognition, Heine (2001) points out some of the important ways in which North Americans and East Asians differ. For instance, there is evidence that East Asians perceive a heightened distinction between ingroups and outgroups. Also, the two cultures seem to differ in who is included in their ingroup circles. If the IAT effects can be conceived as a measure of ingroup-outgroup distinction, the stronger IAT effects displayed by East Asian participants is consistent with this idea. However, the context of this study is very limited, and so we should refrain from drawing conclusions.3 This cultural perspective is still in its infancy, and a great deal of work within this framework is anticipated. With respect to prejudice and stereotyping, theories and hypotheses may need to take into account the cultural background of the individuals involved in order to arrive at a more complete understanding of the phenomena. With respect to gender, Schaller (2001) offers some plausible reasons for supposing that males and females may differ in the activation of the danger-avoidance mechanism. To the extent that males were more often involved in intergroup contact and conflict than females, it is 26 possible that this mechanism either exists solely in males or manifests more strongly in males against other male targets. This is speculative, and the fact that the present study, with mostly female participants, obtained the hypothesized results suggests no reasons to suppose that the mechanism does not exist in females. The present study involved only male targets, and so the effect of target gender could not be examined. Further Comments on the Evolutionary Perspective Any theoretical perspective that illuminates our understanding is valuable. As scientists, we should remain receptive to these useful perspectives. Although the evolutionary perspective is a theoretical perspective no different from others, there has been much resistance towards it. Some of this may arise from the "illusion of unfalsifiability" of evolutionarily derived hypotheses (Schaller & Conway, 2000)—the perception that because the starting assumptions regarding the ancestral environment are speculative, the hypotheses derived from the perspective are unfalsifiable. It should be noted that many theoretical frameworks begin with speculative assumptions, which are only later supported by data. The evolutionary perspective not only generates novel, testable hypotheses, but it may offer a way of carving various psychological processes into meaningful domains (see section below on domain-specificity). Cautionary remarks. As with many evolutionary accounts of psychological processes, there are some cautionary remarks. The relevant points are succinctly summarized by Schaller and Park (in press). First, the fact that certain mechanisms exist in nature has no relation to their moral status. That is, even i f "prejudices have a basis in the natural history of the species, this in no way justifies or excuses prejudicial attitudes and the social problems that they cause" (Schaller & Park, p. 6). Second, an evolutionary basis for prejudice does not suggest that these processes are functional or adaptive in the current environment. (Prejudice, with its obvious detrimental effects, can easily be argued to be maladaptive and nonfunctional.) And third, an evolutionary basis for prejudice does not imply that the process is somehow inevitable and 27 immutable. In fact, as pointed out by Schaller and Park, the evolutionary approach may inform us not only about factors that trigger prejudice, but also those that mitigate it. Domain-specificity. It was mentioned earlier that a major contribution of the evolutionary perspective is the identification of specific mechanisms. Indeed, the danger-avoidance mechanism and the cues that may influence it may not have been arrived at as easily without consideration of the evolutionary history of humans. But the value of this perspective does not end with its generative utility. To the extent that the mind is composed of multiple domain-specific mechanisms (along with a number of domain-general ones), identifying the psychological mechanisms provides us with an accurate taxonomy of the "parts" of the mind (Buss, 1995). The findings of this study demonstrate such domain-specificity. Psychological theories and models often draw somewhat arbitrary boundaries between proposed cognitive components. The evolutionary perspective, with its capacity to identify mechanisms, provides a conceptually coherent organization of the mind, based on function (Buss). The relationship between stereotyping and prejudice. Up to this point, the question of how stereotyping may be related to prejudice has been largely ignored. It is generally agreed that prejudice and stereotyping are inseparable—we cannot examine one without considering the other. Indeed, stereotyping has been conceived as "both a cause and a consequence of prejudice" (Kawakami, Dion, & Dovidio, 1998, p. 407). To the extent that prejudice is viewed as an attitude, stereotypes have been seen as the cognitive component (in the traditional way of viewing attitudes). However, aside from exploring the correlations between stereotype use and prejudice (e.g., Devine, 1989; Kawakami et al., 1998; Lepore & Brown, 1997), most investigators seem to tacitly assume the connection between the two without clearly characterizing the relationship between them. Devine (1989) provides one characterization of this relation by arguing that automatic activation or even conscious knowledge of stereotypes does not imply prejudice against a given 28 group. That is, a prejudiced individual is someone who has knowledge of the stereotypes and personally endorses them. Thus, in her characterization, the relation between stereotyping and prejudice is essentially that of mere knowledge and personal endorsement of that knowledge. Allport (1954), by defining prejudice as "an antipathy based upon a faulty and inflexible generalization" (p. 10), suggests that the direction of causation is from stereotyping to prejudice. That is, we categorize people, form (faulty) beliefs about them, and as a consequence, form negative evaluations about them. However, as noted by Pettigrew (1979), the direction of causation does seem to go the opposite way as well. Incidentally, the evolutionary psychological perspective may provide a way of characterizing the distinction and the connection between stereotyping and prejudice that makes them logically and necessarily inseparable. In plain language, prejudice can be thought of as the part of us that says "I don't like them" and stereotyping can be thought of as the part of us that says "those people are X " (where X is a derogatory trait). So then, it seems reasonable that our minds are saying "I don't like those people precisely because they are X . " Although this link has often been implied, no theory has included a compelling elucidation of this connection. The present study examined outgroup derogation mediated by the danger-avoidance mechanism. To the extent that this mechanism exists, it would make little sense for the mechanism to lead people to form beliefs about outgroup members (e.g., "they are dangerous") without an accompanying affective component to motivate the avoidance (e.g., "I don't like them"). Stereotyping and prejudice, then, become necessarily inseparable. The family of studies to which the current study belongs may have identified one strand of prejudice and stereotyping—the one that says "I don't like them because they are dangerous." Implicit in this conceptualization is that there may be many such strands, strands that are logically independent from one another, and thus, influenced by different factors. Other 29 studies suggest the existence of another mechanism, one that says "I don't like them because they might be diseased." There is some evidence that individual differences in perceived vulnerability to disease (whose effects are moderated by contextual cues connoting disease) are related to prejudice against racial and other stigmatized groups that fit the "diseased other" template (Schaller & Park, in press; Schaller et al., 2002). Importantly, this process appears to be largely unrelated to the danger-avoidance mechanism. Practical Implications Research of stereotyping processes would seem fruitless i f the findings did not suggest ways of preventing or mitigating stereotyping and related outgroup derogation processes. An obvious implication is to be wary of the effects of ambient darkness and chronic beliefs about danger. Good lighting during nighttime hours is known to reduce criminal activity (Quinet & Nunn, 1998), and the results from the present study add one more reason to have good lighting in certain situations. A possible intervention strategy might include reducing chronic beliefs about danger or promoting feelings of safety. Mikulincer and Shaver (2001) provide evidence that subliminally priming words connoting safety may reduce prejudiced attitudes toward potential immigrants. Also, Fein and Spencer (1997) showed that individuals whose self-images were bolstered through self-affirmation procedures were less likely to view a stereotyped group negatively. On a broader level, i f processes underlying prejudice and stereotyping are a collection of multiple, distinct processes, then intervention tactics may need to be just as specific as well (Schaller & Park, in press). For instance, promoting feelings of safety and security may be useful for mitigating racism, but not sexism. Future Directions and Concluding Remarks Intergroup vigilance theory offers additional hypotheses that can be explored. The effects of B D W and ambient darkness may extend to other intergroup cognitive processes, such 30 as impression formation and person memory of ingroup and outgroup members, especially when danger-relevant traits are involved (Schaller, 2001). In addition, there may be other heuristic cues analogous to ambient darkness—such as outgroup size and outgroup proximity— that also connote danger, and so influence the danger-avoidance mechanism. It may also be useful to explore the impact of culture and gender differences, i f any. It may also be useful to search for processes that parallel the danger-avoidance mechanism (such as the disease-avoidance mechanism). Perception of danger appears to be an important causal factor of racial stereotyping, but it is only one factor. In searching for the roots of prejudice and stereotyping, a pluralistic approach may be most fruitful. To the extent that prejudice is multifaceted, looking for a few, grand theories may prevent us from identifying the intricacies of the phenomenon. 31 References Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N . (1950). The authoritarian personality. New York: Harpers. Allport, G. (1954). The nature of prejudice. New York: Addison-Wesley. Altemeyer, B. (1988). Enemies of freedom. San Francisco: Jossey-Bass. Ashburn-Nardo, L. , Voils, C. I., & Monteith, M . J. (2001). Implicit associations as the seeds of intergroup bias: How easily do the take root? Journal of Personality and Social Psychology, 81, 789-799. Banaji, M . R., & Hardin, C. D. (1996). Automatic stereotyping. Psychological Science, 7, 136-141. Blair, I. V. , & Banaji, M . R. (1996). Automatic and controlled processes in stereotype priming. Journal of Personality and Social Psychology, 70, 1142-1163. Blair, I. V. , Ma, J. E., & Lenton, A . P. (2001). Imagining stereotypes away: The moderation of implicit stereotypes through mental imagery. Journal of Personality and Social Psychology, 81, 828-841. Brendl, C. M . , Markman, A . B. , & Messner, C. (2001). How do indirect measures of evaluation work? Evaluating the inference of prejudice in the implicit association test. Journal of Personality and Social Psychology, 81, 760-773. Brewer, M . B. (1979). In-group bias in the minimal intergroup situation: A cognitive-motivational analysis. Psychological Bulletin, 86, 307-324. Buss, D. M . (1995). Evolutionary psychology: A new paradigm for psychological science. Psychological Inquiry, 6, 1-30. Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116-131. Campbell, D. T. (1965). Ethnocentric and other altruistic motives. In D. Levine (Ed.), Nebraska Symposium on Motivation (pp.283-311). Lincoln: University of Nebraska Press. Dasgupta, N . , & Greenwald, A. G. (2001). On the malleability of automatic attitudes: Combating automatic prejudice with images of admired and disliked individuals. Journal of Personality and Social Psychology, 81, 800-814. Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Journal of Personality and Social Psychology, 56, 5-18. Fazio, R. FL, Jackson, J. R., Dunton, B. C , & Williams, C. J. (1995). Variability in automatic activation as an unobtrusive measure of racial attitudes: A bona fide pipeline? Journal of Personality and Social Psychology, 69, 1013-1027. Fein, S., & Spencer, S. J. (1997). Prejudice as self-image maintenance: Affirming the self through derogating others. Journal of Personality and Social Psychology, 73, 31-44. Fiske, S. T. (1998). Stereotyping, prejudice, and discrimination. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology, Vol. 2 (4th ed.) (pp. 357-411). New York: McGraw-Hill. Greenwald, A . G., McGhee, D. E., & Schwartz, J. L. K . (1998). Measuring individual differences in implicit cognition: The implicit association test. Journal of Personality and Social Psychology, 74, 1464-1480. Grillon, C , Pellowski, M . , Merikangas, K . R., & Davis, M . (1997). Darkness facilitates acoustic startle reflex in humans. Biological Psychiatry, 42, 453-460. Heine, S. J. (2001). Self as cultural product: An examination of East Asian and North American selves. Journal of Personality, 69, 881-906. Karpinski, A. , & Hilton, J. L. (2001). Attitudes and the implicit association test. Journal of Personality and Social Psychology, 81, 774-788. Kawakami, K. , Dion, K. L. , & Dovidio, J. F. (1998). Racial prejudice and stereotype activation. Personality and Social Psychology Bulletin, 24, 407-416. Lepore, L. , & Brown, R. (1997). Category and stereotype activation: Is prejudice inevitable? Journal of Personality and Social Psychology, 72, 275-287. Lowery, B. S., Hardin, C. D., & Sinclair, S. (2001). Social influence effects on automatic racial prejudice. Journal of Personality and Social Psychology, 81, 842-855. Mikulincer, M . , & Shaver, P. R. (2001). Attachment theory and intergroup bias: Evidence that priming the secure schema attenuates negative reactions to out-groups. Journal of Personality and Social Psychology, 81, 97-115. Neuberg, S. L. , & Newsom, J. T. (1993). Personal need for structure: Individual differences in the desire for simple structure. Journal of Personality and Social Psychology, 65, 113-131. Page, R. A. , & Moss, M . K. (1976). Environmental influences on aggression: The effects of darkness and proximity of victim. Journal of Applied Social Psychology, 6, 126-133. Pettigrew, T. F. (1979). The ultimate attribution error: Extending Allport's cognitive analysis of prejudice. Personality and Social Psychology Bulletin, 5, 461-476. Quinet, K . D., & Nunn, S. (1998). Illuminating crime: The impact of street lighting on calls for police service. Evaluation Review, 22, 751-779. Rudman, L. A. , Greenwald, A. G., Mellott, D. S., & Schwartz, J. L. K . (1999). Measuring the automatic components of prejudice: Flexibility and generality of the implicit association test. Social Cognition, 17, 437-465. Schaller, M . (2001). Intergroup vigilance theory. Unpublished Manuscript, University of British Columbia. 34 Schaller, M . (in press). Ancestral environments and "motivated" social perception: Goal-like blasts from the evolutionary past. In S. Spencer & S. Fein (Eds.), Motivated social cognition: The Ninth Ontario Symposium. Mahwah, NJ: Erlbaum. Schaller, M . , & Conway, L. G., III. (2000). The illusion of unfalsifiability and why it matters. Psychological Inquiry, 11, 49-52. Schaller, M . , Kelly, C , & Dhanoa, G. (2001). Threat and prejudice: Effects of chronic and temporary indicators of physical danger on attitudes toward immigrants and other 'outsiders,' Unpublished manuscript, University of British Columbia. Schaller, M . , & Park, J. H . (in press). Prehistoric dangers and contemporary prejudices. European Review of Social Psychology. Schaller, M . , Park, J. PL, Faulkner, J., Duncan, L. A. , Kelly, C , & Dhanoa, G. (2002). Perceived vulnerability to disease and its relation to ethnic prejudice. Unpublished manuscript, University of British Columbia. Sherif, M . , Harvey, O. J., White, B. J., Hood, W. R., & Sherif, C. W. (1961). Intergroup conflict and co-operation: The Robber's Cave experiment. Norman: University of Oklahoma. Spencer, S. J., Fein, S., Wolfe, C. T., Fong, C , & Dunn, M . A . (1998). Automatic activation of stereotypes: The role of self-image threat. Personality and Social Psychology Bulletin, 24, 1139-1152. Stephan, W. G., Ybarra, O., Martinez, C. M . , Schwarzwald, J., & Tur-Kaspa, M . (1998). Prejudice toward immigrants to Spain and Israel: An integrated threat theory analysis. Journal of Cross-Cultural Psychology, 29, 559-576. Tooby, J., & Cosmides, L. (1989). Evolutionary psychology and the generation of culture, Part I. Theoretical considerations. Ethology and Sociobiology, 10, 29-49. Wittenbrink, B., Judd, C. M . , & Park, B. (1997). Evidence for racial prejudice at the implicit level and its relationship with questionnaire measures. Journal of Personality and Social Psychology, 72, 262-21A. Wittenbrink, B. , Judd, C. M . , & Park, B. (2001). Spontaneous prejudice in context: Variability in automatically activated attitudes. Journal of Personality and Social Psychology, 81, 815-827. 36 Footnotes 'it is worth noting that, for the most part, the results of previous studies revealed only small and non-significant order effects. 2It was initially anticipated that asking participants to think about African Americans would lead to a greater likelihood of observing noticeable effects. This part of the procedure was modified when the results obtained from the first group of participants did not reveal any interpretable trends. 3Steven Heine suggested a simpler alternative: Minority members (East Asians in this context) may be endorsing the ingroup to a greater extent than majority group members. 37 Table 1 Correlations among the Individual Difference Measures Variable 1 2 3 4 5 1. PNS (.77) -.18 .31* .09 .17 2. NFC (.87) .14 -.09 -.16 3. MCPR (.79) .22 .12 4. BDW (.87) .31* 5. PVD (.79) Note. Internal consistency (Cronbach's alpha) of each scale is on the diagonal. *p<.05. The abbreviations denote: personal need for structure (PNS), need for cognition (NFC), motivation to control prejudiced responding (MCPR), belief in a dangerous world (BDW), and perceived vulnerability to disease (PVD). Table 2 Mean IAT Effects of High and Low BDW Participants in Dark and Light Conditions Low BDW High BDW Light Dark Light Dark Unpleasant-pleasant IAT 153.74 66.25 141.31 125.25 Danger-safety IAT 226.45 20.39 137.69 179.44 Table 3 ANOVA Results on Effects of Darkness and BDW on the Unpleasant-Pleasant and Danger-Safety IAT Effects Unpleasant-Pleasant IAT Effect Source df MS F p Darkness 1 33736.57 .62 .44 BDW 1 6823.50 .13 .73 Darkness X BDW 1 16052.66 .30 .59 Error 47 54467.34 Danger-Safety IAT Effect Source df MS F p Darkness 1 84946.31 2.65 .11 BDW 1 15549.35 .485 .49 Darkness X BDW 1 193207.21 6.03 .02 Error 47 32046.94 40 Table 4 Results of Regression Analyses on Effects of Darkness and BDW on the Unpleasant-Pleasant and Danger-Safety IAT Effects Unpleasant-Pleasant IAT: Implicit Associations of Africans with Unpleasant Predictor P Darkness main effect 25.85 .11 .79 .44 BDW main effect 18.15 .08 .55 .59 Darkness X BDW interaction -34.51 -.15 -1.03 .31 Danger-Safety IAT: Implicit Associations of Africans with Danger Predictor B Darkness main effect 37.29 .20 1.44 .16 BDW main effect 24.38 .13 .94 .36 Darkness X BDW interaction -49.98 -1.89 -1.89 .07 41 Block 1 2 3 4 5 Task word face combined reversed reversed description categori- categori- task face combined zation zation categori- task zation • African • Asian Reminders • unpleasant • African • unpleasant • Asian • unpleasant pleasant • Asian • Asian • African • African • pleasant • pleasant • ° dirty Asian ° ° dirty African ° ° Asian Sample intelligent ° ° African Asian - African ° intelligent ° Stimuli ° failure Asian ° laughter ° ° Asian African -(faces were ° lazy Asian - ° African African ° ° dirty actual laughter ° ° African Asian ° ° Asian ° Asian photos) ° poverty Asian ° ° poverty African ° pleasure ° pleasure ° - African rainbow ° African ° African ° rainbow ° - African ° African ° Asian ° failure F/gure 1. Schematic illustration of the implicit association test (IAT): An example of the unpleasant-pleasant IAT taken by an Asian participant. Figure 2. Unpleasant-pleasant IAT effects among high- and low-BDW participants the Light and Dark conditions. 43 Figure 3. Danger-safety IAT effects among high- and low-BDW participants in the Light and Dark conditions. Figure 4. Correlation between BDW and unpleasant-pleasant IAT effect. Figure 5. Correlation between BDW and danger-safety IAT effect. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items