Summarizing a Published Work
Cultural Stereotypes and the Self: A Closer Examination of Implicit Self-Stereotyping
Janetta Lun
University of Virginia
Stacey Sinclair
Princeton University
Courtney Cogburn
University of Michigan
Recent research and theory on implicit self-stereotyping suggests that individuals nonconsciously incorporate stereotypes about their social groups into the self-concept; however, evidence as to whether this holds true for negative stereotypes remains limited. Using a subliminal priming measure, the current research found that women (Experi- ment 1) and White Americans (Experiment 2) implicitly associated the self with in-group stereotypic traits but not out-group stereotypic traits. Of importance, both groups implicitly self-stereotyped on negative in-group traits to a similar extent as they did on positive in-group traits. Moreover, exploratory analysis showed that the degree to which White Americans associated positive, but not negative, in-group stereotypes with the self was related to higher self-esteem. Implications of implicit self-stereotyping on self-esteem and stereotype-consistent behavior are discussed.
Cultural stereotypes are widely known beliefs (Devine, 1989; Devine & Elliot, 1995; Katz & Braly, 1933) that broadly influence how individuals are evaluated and treated (Fiske, 1998; Hamilton, Sherman, & Ruvolo, 1990; Hilton & von Hippel, 1996; Macrae & Bodenhausen, 2000). Because of the pervasiveness of these beliefs, targets of stereotypes live in a world where they are frequently assumed to have stereotypic traits or behave in a stereotype-consistent manner (Bargh & Pietromonaco, 1982; Darley & Gross, 1983; Devine, 1989; Fiske, Lin, & Neuberg, 1999; Srull & Wyer, 1979). Moreover, people who hold stereotypic expecta- tions can elicit behavior from targets that confirms these expectancies (e.g., Olivier & Snyder, 2003; Word, Zanna, & Cooper, 1974).
Influential theory and research suggests that continuous exposure to, being evaluated in terms of, and occasionally behaving in a manner consistent with stereotypes may lead targets to incorporate stereotypic beliefs about their social groups into their own self- concept (Allport, 1954=1979; Cooley, 1902; Mead, 1934; Tice & Wallace, 2003), even those aspects of the self-concept that are less available for conscious intro- spection (i.e., implicit; Greenwald & Banaji, 1995; see Devos & Banaji, 2003, for a review). We refer to the nonconscious incorporation of in-group stereotypes into the self-concept as implicit self-stereotyping.
The unified theory of implicit cognition, a prominent theoretical explanation of implicit self-stereotyping, sug- gests that members of a given social group, on average, ought to implicitly associate stereotypes of their group with the self (Greenwald et al., 2002). According to this perspective, individuals will implicitly associate in-group stereotypes with the self to the extent that they also implicitly associate the self with the group identity and
Correspondence should be sent to Janetta Lun, 102 Gilmer Hall,
P.O. Box 400400, Charlottesville, VA 22904–4400. E-mail: jlun@
virginia.edu
BASIC AND APPLIED SOCIAL PSYCHOLOGY, 31:117–127, 2009
Copyright # Taylor & Francis Group, LLC
ISSN: 0197-3533 print=1532-4834 online
DOI: 10.1080/01973530902880340
the group with in-group stereotypes. Given that people tend to associate the self with the in-group more so than with out-groups (Nosek, Banaji, & Greenwald, 2002b; see Greenwald et al., 2002, for a review) and are as likely to hold implicit stereotypes of in-groups as out-groups (Nosek, Banaji, & Greenwald, 2002a, 2002b), it follows that members of a social group should, on average, implicitly associate in-group stereotypes with the self more so than out-group stereotypes (Greenwald & Farnham, 2000; Smith & Henry, 1996). Consistent with this logic, research has shown that women as a group are more likely to associate positive feminine traits (e.g., gentle, warm) with their self-concepts than men do (Farnham & Greenwald, 2000). Based on the stereotype that men are more powerful and women are weaker, Rudman, Greenwald, and McGhee (2001) also showed that men associate attributes related to power, relative to those related to being weak, with their self-concepts to a greater extent than women do. This evidence suggests that people do implicitly associate in-group stereotypes with the self.
Although extant research has demonstrated implicit self-stereotyping, the limits of this phenomenon remain unexamined. Do people, on average, implicitly associate positive and negative in-group stereotypes with the self to an equal degree? Or, do people selectively associate one type of in-group stereotype with the self more than another? These questions call for empirical attention because different theoretical perspectives suggest differ- ent answers.
On one hand, some theory and research suggests that people would implicitly self-stereotype with respect to positive but not negative in-group stereotypic traits. According to social identity theory, people tend to seek positive self-views through their social group member- ships (Tajfel & Turner, 1986). Although this theory does not explicitly take a stand with respect to unconscious thought processes, it is not unreasonable to assume that the motivation to sustain positive self-views through group memberships operates automatically (Greenwald et al., 2002). In fact, in-group favoritism, the tendency for people to evaluate their in-group more favorably than out-groups to enhance positive self-esteem, has been shown to occur automatically (see Dasgupta, 2004, for a review). For example, people have more automatic positive evaluations toward in-group designa- ting pronouns (e.g., us, we) than outgroup designating pronouns (e.g., they, them; Perdue, Dovidio, Gurtman, & Tyler, 1990). Research has also demonstrated that implicit positive evaluation toward the in-group among high-status group members is associated with higher implicit self-esteem (Jost, Pelham, & Carvallo, 2002).
The unified theory of implicit cognition can also be construed as predicting self-stereotyping on positive but not negative in-group traits. This perspective assumes
that the self is characterized by positive traits, that is, the core self tends to be associated with more positive traits than negative traits (Greenwald et al., 2002; Rudman et al., 2001). As such, people should be more apt to associate positive in-group stereotypic traits with the self than negative in-group stereotypic traits. How- ever, research generated by this perspective has yet to test this prediction in a way that unconfounds stereotype content and valence or that directly compares implicit self-stereotyping on positive versus negative stereotypes.
On the other hand, several lines of thinking suggest that individuals should associate both negative and posi- tive stereotypic attributes with the self (e.g., Allport, 1954=1979; Cooley, 1902; Jost, Banaji, & Nosek, 2004; Mead, 1934; Rudman, 2004). First, this outcome is implied by the notion that self-stereotyping results from the knowledge of, and continuous exposure to, in-group stereotypes. After all, most widely recognized stereo- types have both negative and positive aspects (Fiske, Cuddy, Glick, & Xu, 2002), and members of stereotyped groups are well aware of both the positive and negative beliefs about their groups (Crocker, Major, & Steele, 1998; Devine & Elliot, 1995). Moreover, one could argue that implicit measures of the self-concept are more likely than explicit measures to reveal associations between negative stereotypes and the self because parti- cipants are less likely to be aware that their self-views are being assessed and, even if they are aware, implicit measures are less subject to strategic control (cf. Dasgupta, McGhee, Greenwald, & Banaji, 2000; Greenwald, McGhee, & Schwartz, 1998).
Self-categorization theory also suggests that people should associate both positive and negative in-group traits to the self (e.g. Simon, Glassner-Bayerl, & Stratenwerth, 1991; Simon & Hamilton, 1994; Smith, Coats, & Walling, 1999; Smith & Henry, 1996; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). In support of this perspective, for example, Smith and Henry found that individuals’ self-concepts are more closely associa- ted with both the positive and negative traits they think characterize the social group that they belong to. Although they found the same pattern of results with consensual stereotypes of the group, it was unclear whe- ther these stereotypes included both positive and negative traits. Therefore, a stronger case can be made by research that directly compares positive and negative culturally shared stereotypes, as we do in the current research.
Finally, system justification theory contends that positive and negative stereotypes serve to justify the advantages and disadvantages that people are subject to (Jost, 2001; Jost & Banaji, 1994). Over time, justi- fication process may lead to incorporation of these stereotypes into the self-concept (Dunham, Baron, & Banaji, 2006). Although studies of system justification do not directly assess such self-stereotyping, indirect
118 LUN, SINCLAIR, COGBURN
support for this contention comes from research showing that members of low-status groups tend to favor the out-group over their own group, whereas members of high-status groups show in-group favoritism (Jost & Burgess, 2000; Mullen, Brown, & Smith, 1992; Rudman, Feinberg, & Fairchild, 2002). Such out-group favoritism is particularly evident when assessed using implicit mea- sures (Jost et al., 2004). As targets use stereotypes to jus- tify their in-group status, they may subtly reinforce the association between these stereotypes and the in-group or the self. More important, stereotypes and beliefs do not have to be of a particular valence to justify the status of a group. For example, the stereotype that poor people are happier and more honest than rich people serves to justify the inferior status of the poor (Kay & Jost, 2003). Thus, this theory and related evidence concurs that peo- ple should associate both positive and negative in-group stereotypes with the self—particularly on an implicit measure.
Our research speaks to the opposing predictions by examining implicit self-stereotyping with people from two social groups that are known to be associated with positive and negative stereotypes: women and White Americans. We first examined implicit self-stereotyping among women (Experiment 1). Previous research has shown that women are more likely to associate feminine traits with the self than masculine traits (Greenwald & Farnham, 2000), but the self-stereotyping effect was confounded with the positive nature of feminine traits and the study did not specifically examine trait valence. In our study, we predict that women will implicitly associate positive female stereotypes with the self, pro- viding evidence convergent with previous research. The critical test then is whether women will implicitly incor- porate negative female stereotypic traits into the self-concept. As previously mentioned, there are theore- tical reasons to expect or to not expect they would do so.
Then we extended our examination to White Ameri- cans and tested whether members of this group engaged in implicit self-stereotyping (Experiment 2). Although members of this group are less likely to be thought of as targets of stereotypes, they are not exceptions to the socialization of cultural stereotypes (Rudman, 2004). Thus, White Americans may be as likely as other social groups to implicitly associate in-group stereotypes with the self, thus providing strong evidence of implicit self-stereotyping. For exploratory purposes, we also examined whether implicit self-stereotyping among White Americans is related to their self-esteem.
EXPERIMENT 1
This experiment examined whether women implicitly associated positive and negative feminine traits with
the self. Unlike past research, we measured implicit self-stereotyping with a sequential subliminal priming task as opposed to the Implicit Association Test (IAT; Greenwald et al., 1998) to capture the association between the self and different stereotypic traits. There are two advantages to using a priming task instead of the IAT for the purpose of this research. First, it addresses the limitation of relative difference inherent in the IAT (Nosek & Banaji, 2001). The gender self-stereotyping IAT shows, for example, how much women are likely to associate feminine traits with the self as compared to masculine traits. Findings using this measure are not able to discern whether women have a stronger association between feminine traits and the self or a weaker association between masculine traits and the self. Second, when using a subliminal priming paradigm, people will be completely unaware of the relevance of stereotypes to this task, thereby minimizing positive self-presentation or reactance effects.
Method
Participants
Thirty-eight female undergraduates participated in this experiment to partially fulfill a course requirement. They all identified themselves as White.
Procedure
Participants were recruited to a study ostensibly about word identification. Upon arriving at the labora- tory, one of the two female White experimenters gave participants instructions for a computerized lexical deci- sion task that followed the procedure of the subliminal priming task in Wittenbrink, Judd, and Park (1997). They were told the following: Their goal was to distin- guish, as quickly as they could, whether a series of letter strings presented on the computer were words or non- words. To help them respond quickly, they were to focus on a black dot at the center of the screen throughout the task, where the letter strings would appear, and press the F key marked ‘‘W’’ when they thought the presented let- ter string was a word but should press the J key marked ‘‘N’’ when they thought the letter string was a nonword. To ensure that participants understood the instructions, the experimenter stayed with them through the practice trials and answered any questions they had. Then parti- cipants were left alone to complete the experimental trials. After the lexical decision task, the experimenter returned with a self-evaluation questionnaire for partici- pants to complete. Although this was not the focus of the study, we used this questionnaire to measure the degree to which participants consciously ascribed stereo- typic traits to the self. Finally, participants were probed for suspicion and whether they detected any of the prime
IMPLICIT SELF-STEREOTYPING 119
words in the lexical decision task. After that, they were debriefed and thanked.
Materials
Lexical decision task. The computer task was used to assess the degree to which participants implicitly associated female and male stereotypes with the self. In a series of trials, self-relevant or neutral words were subliminally presented prior to presentations of a female stereotypic word, male stereotypic word, or a nonword that participants judged as a word or nonword. The dependent measure was participant’s response latencies when making these lexical judgments (e.g., Wittenbrink et al., 1997).
The words in this task were positive and negative female and male stereotypes. The positive female stereo- types were caring, compassionate, faithful, and sensitive, and the negative female stereotypes were complaining, dependent, moody, shy, and weak. The positive male stereotypes were athletic, confident, powerful, and strong, and the negative male stereotypes were aggressive, arro- gant, insensitive, selfish, and stubborn. To show that these traits were valid gender stereotypes in the specific population examined, a separate group of 36 undergrad- uate students rated the extent to which each of these traits was perceived to be mostly associated with women versus men on a 7-point Likert-type scale from 1 (mostly associated with women) to 7 (mostly associated with men). All of the female stereotypic traits were rated signifi- cantly below the midpoint (4) and all of the male stereo- typic traits were rated above the midpoint (all ps < .05). Another group of 32 students rated the extent to which each of these traits was perceived as negative versus positive on a 1 (very negative) to 7 (very positive) Likert-type scale. As expected, all of the positive traits were rated significantly above the neutral point, whereas all of the negative traits were rated significantly below the neutral point (all ps < .05). Twenty-two participants made lexical judgments on the 18 stereotypic words and 10 nonwords twice in randomized order for a total of 56 trials. Five participants completed this task making judgments on these words and nonwords only once for a total of 28 trials. The number of trials did not affect the results when it was entered as a factor, and therefore the data were combined in the analyses.
For each trial, participants first saw a black dot, a fixation point, in the middle of the screen that they were told to focus on. At the same time, a string of Xs pre- sented at one of the corners of the screen acted as a forward mask of the prime. After 1,000 msec, a self (self, I, me) or neutral word (a, at, the) was presented sublim- inally for 15 msec in place of the mask. It was followed by the same string of Xs, which acted as the backward mask for the prime for another 1,000 msec. Immediately
after the presentation of the prime and masks, a female stereotypic word, male stereotypic word, or nonword appeared in place of the fixation point. Participants then made the word or nonword judgment, and the latencies of these judgments were recorded. If participants responded incorrectly, the program waited for the correct response before proceeding to the following trial. Partici- pants first completed four practice trials including two words (i.e., apple, pencil) and two nonwords (i.e., youey, njoue) to become familiarized with the keys. Then they pro- ceeded to the experimental trials. Implicit self-stereotyping was indicated by how much the self primes, compared to the neutral primes, facilitated individuals’ response latency to female and male stereotypic words.
Questionnaire. After the lexical decision task, parti- cipants completed a questionnaire that assessed their explicit self-stereotyping on the stereotypic traits that were presented in the lexical decision task, as well as some filler traits. They were asked to rate the degree to which each of these traits applied to themselves on a 7-point Likert-type scale from 1 (not at all) to 7 (very much). Four composites were created to reflect the extent to which participants thought each type of stereo- typic traits applied to them: average selfratings on positive female stereotypes (a ¼ .68), negative female stereotypes (a ¼ .67), positive male stereotypes (a ¼ .70), and negative male stereotypes (a ¼ .80).
Results
Two participants were excluded from the analyses because they reported seeing a self-related word between the masks. Although three participants reported seeing flashes in the masks, they could not identify any of the actual prime words so they were not excluded from the following analyses. Results with or without these three participants did not differ.
Several procedures were performed to normalize the distribution of response latencies. First, we eliminated outliers using criteria adapted from previous research (Bargh & Chartrand, 2000; Fazio, 1990; Wittenbrink et al., 1997). Latencies that were more than 3 standard deviations above individuals’ average reaction time, lower than 150 msec, and from errors that were later corrected were recorded as missing values. A total of 20 latencies were recorded as missing values, 1.6% of a total of 1,188 word trials. We then log transformed the latencies (Fazio, 1990; Ratcliff, 1993). For ease of interpretation, descriptive statistics are reported in milliseconds.
Implicit Self-Stereotyping
If women are subject to implicit self-stereotyping, they should respond to female stereotypes faster after
120 LUN, SINCLAIR, COGBURN
self-relevant primes than neutral primes. In addition, they should respond equally fast, or slower, to male stereotypes after self-relevant primes versus neutral primes. To test this hypothesis, the transformed laten- cies were entered into a 2 (prime: self vs. neutral primes) � 2 (traits: female vs. male stereotypes) � 2 (valence: positive vs. negative words) repeated measure analysis of variance (ANOVA). In general, participants responded faster to positive words (M ¼ 681, SD ¼ 262) than negative words (M ¼ 715, SD ¼ 255), F(1, 35) ¼ 10.25, p < .01, partial g2 ¼ .23. They also responded fas- ter to male stereotypes (M ¼ 684, SD ¼ 253) than female stereotypes (M ¼ 713, SD ¼ 266), F(1, 35) ¼ 6.27, p ¼ .02, partial g2 ¼ .15.
As hypothesized, there was also an interaction between prime and traits, F(1, 35) ¼ 5.82, p ¼ .02, partial g2 ¼ .14 (see Figure 1). Simple effects analysis revealed that participants responded faster to female stereotypes following selfprimes (M ¼ 688, SD ¼ 252) than to the same stereotypes following neutral primes (M ¼ 739, SD ¼ 298), t(35) ¼ 2.00, p ¼ .05, partial g2 ¼ .10. How- ever, there was no difference in the speed with which they responded to male stereotypes following self-primes (M ¼ 697, SD ¼ 281) than to the same stereotypes fol- lowing neutral primes (M ¼ 672, SD ¼ 242), t(35) < 1, p > .30. This pattern was not moderated by valence, F(1, 35) ¼ 1.73, p ¼ .20.
Explicit Self-Stereotyping
Similar to the analysis of implicit self-stereotyping, participants’ selfratings on stereotypic traits were sub- mitted to a 2 (traits: female vs. male stereotypes) � 2 (valence: positive vs. negative words) repeated measure ANOVA. Participants rated themselves higher on female stereotypic traits (M ¼ 4.84, SD ¼ 0.56) than male stereotypic traits (M ¼ 4.16, SD ¼ 0.73), F(1, 35) ¼ 33.75, p < .01, partial to g2 ¼ .49, and higher on
positive traits (M ¼ 5.29, SD ¼ 0.63) than negative traits (M ¼ 3.71, SD ¼ 0.89), F(1, 35) ¼ 82.22, p < .01, partial to g2 ¼ .70. Moreover, these main effects were qualified by an interaction between traits and valence, F(1, 35) ¼ 28.12, p < .01, partial to g2 ¼ .45. Female participants rated themselves as more stereotypically feminine (M ¼ 5.95, SD ¼ 0.61) than masculine (M ¼ 4.71, SD ¼ 0.90) on positive traits, t(35) ¼ 8.07, p < .01, partial to g2 ¼ .65, but this difference was attenuated on negative traits (Mf ¼ 3.70, SD ¼ 0.94 vs. Mm ¼ 3.44, SD ¼ 1.10), t(35) ¼ 1.59, p ¼ .12.
Positive and Negative Self-Stereotyping on Implicit Versus Explicit Measures
To directly compare the valence differences in self-stereotyping on the implicit and explicit measure on the same metric, we converted the degree to which women implicitly and explicitly self-stereotyped on positive versus negative traits to Cohen’s ds.1 For expli- cit self-stereotyping, we first computed two difference scores that captured the degree to which participants explicitly self-stereotyped on positive and negative traits (self-ratings on positive female stereotypes – self-ratings on positive male stereotypes; self-ratings on negative female stereotype – self-ratings on negative male stereo- types). Then, we derived the Cohen’s d from the mean difference and pooled standard deviation of these two difference scores.
For implicit self-stereotyping, we took an additional step prior to computing the two different scores of implicit self-stereotyping on positive and negative traits because of the task design. We first computed four scores that captured the degree to which participants implicitly associated positive and negative, female and male stereotypes to the self. These scores were created by subtracting latencies in response to a particular type of stereotypes preceded by self primes from the latencies to the same set of traits preceded by neutral primes. After obtaining these four scores, we followed the same procedure as we did with the explicit measure. Another two difference scores were created to indicate the degree to which participants implicitly self-stereotyped posi- tively and negatively (e.g., implicit self-stereotyping on positive female stereotypes – implicit self-stereotyping on positive male stereotypes). Finally, we derived the Cohen’s d based on the mean difference and the pooled standard deviation of these differences. A positive value of a d score indicated positive self-stereotyping and a negative value suggested negative self-stereotyping.
On the explicit measure, participants’ self-ratings on positive in-group traits, on average, were nearly 1 stan- dard deviation above their ratings on negative in-group
FIGURE 1 Average response latencies of women to male and female
stereotypes after self versus neutral primes.
1We thank a reviewer for this suggestion.
IMPLICIT SELF-STEREOTYPING 121
traits after controlling for their self-ratings on out-group stereotypes (d ¼ .90), and the d score was reliably different from 0, t(35) ¼ 5.30, p < .001. However, on the implicit measure, the favor toward positive self- stereotyping virtually disappeared (d ¼ �.28). In fact, it was trending in the opposite direction indicating a small propensity to implicitly associate negative in-group traits with the self, but the d score was not reliably different from 0, t(35) ¼ 1.71, p ¼ .10. A paired sample t test further showed that women were more likely to positively self-stereotype on the explicit than implicit measure, t(35) ¼ 5.19, p < .001. These findings are consistent with the results we found previously.
Relationship between Explicit and Implicit Self-Stereotyping
As mentioned, individual differences in the extent to which people implicitly associated the four types of traits (i.e., positive female and male stereotypes, negative female and male stereotypes) with the self-concept were indicated by the difference in latencies for a given type of stereotypic traits preceded by self primes and latencies for the same type of traits preceded by neutral primes. Thus, higher values indicated greater implicit association between that type of trait and the self. These scores were then correlated with explicit self-ratings on the same traits. No consistent relationship was found between implicit and explicit measures of self-stereotyping with respect to a given type of trait. Correlations coefficients for female positive, female negative, male positive and male negative stereotypic traits were r(36) ¼ �.27, .29, .05, and .26 respectively, all ps > .09.
Discussion
Consistent with previous research, Experiment 1 demonstrated that women implicitly associated female stereotypes with the self more strongly than male stereo- types. Of importance, this occurred for both positive and negative in-group stereotypic traits. In contrast, women explicitly self-stereotyped more so on positive than negative traits. Although we found that women self-stereotyped implicitly and explicitly, there was no consistent relationship between implicit and explicit self-stereotyping.
EXPERIMENT 2
In Experiment 2, we examined implicit self-stereotyping among White Americans. Because White Americans are a group that is not commonly identified as the target of stereotypes in part because of their majority status rela- tive to other ethnic groups, simply determining whether they implicitly self-stereotype with respect to this social
identity is illuminating. More important, should they implicitly self-stereotype, it is interesting to examine whe- ther White Americans implicitly self-stereotype with negative in-group stereotypic traits to the same extent as positive in-group stereotypic traits. Finally, we explored whether the degree to which White Americans associated these traits with the self was related to self-esteem in addi- tion to the explicit self-stereotyping measure. Because we did not find a consistent correspondence between implicit and explicit self-stereotyping in Experiment 1, we want to explore whether implicit self-stereotyping will manifest in other explicit judgments about the self such as one’s global or state self-esteem.
Method
Participants
Twenty-eight White undergraduate students (10 female and 18 male) participated in this experiment to partially fulfill a course requirement.
Materials and Procedure
The procedure was the same as in Experiment 1 with one exception. The stereotype-relevant words partici- pants saw in the lexical decision task and the question- naire were stereotypic traits associated with African Americans and White Americans. The African American positive stereotypes used in this study were religious, funny, strong, and athletic, and negative stereotypes of this group were dangerous, violent, loud, uneducated, and poor. The White positive stereotypes were success- ful, rich, wealthy, educated, and preppy, and negative stereotypes of this group were materialistic, stuck-up, racist, and snobby. Thirty-six undergraduate students rated the extent to which each of these traits was perceived by most people as associated with African Americans versus Whites on a 7-point Likert-type scale from 1 (mostly associated with African Americans) to 7 (mostly associated with Whites). Average ratings were all significantly different from the midpoint (4) in the expected direction (all ps < .05). Another group of 32 students rated the extent to which each of these traits was perceived as negative versus positive on a 1 (very negative) to 7 (very positive) Likert-type scale. All of the positive traits, except for preppy, were rated signifi- cantly above the neutral point (4), whereas all of the negative traits were rated significantly below the mid- point (all ps < .01). Preppy was categorized as a positive trait in the analyses reported below to allow a conserva- tive test of negative implicit self-stereotyping.2 The 18
2 The reported pattern data was the same regardless of whether the
trait preppy was coded as a positive trait, coded as a negative trait, or
excluded.
122 LUN, SINCLAIR, COGBURN
stereotypic words and ten nonwords were presented once for half of the participants for a total of 28 trials and twice for the other half of the participants for a total of 56 trials. The number of trials did not affect the results when entered as a factor, and therefore the data were combined in the analyses.
After the computer task, participants explicitly rated how much the stereotypic traits they saw in the computer task, as well as filler items, described them on a 7-point Likert-type scale from 1 (not at all) to 7 (very much). Self-ratings on the stereotypic traits were averaged to produce four scores: positive and negative stereotypes of Whites (a ¼ .68 and a ¼ .72, respectively), and positive and negative stereotypes of African Americans (a ¼ .54 and a ¼ .47, respectively). Finally, participants also com- pleted a global self-esteem scale (Rosenberg, 1965) and a state self-esteem scale (Heatherton & Polivy, 1991; Social and Performance subscales) to examine whether implicit self-stereotyping related to self-relevant judgments. Both scales were reverse-scored and averaged across items, respectively, with higher numbers indicating greater global and state self-esteem.
Results and Discussion
The procedures for eliminating outliers and transforming the latency data to normalize the distribution were the same as in Experiment 1. Eighteen missing values were recorded, 2.2% of a total of 810 word trials. Four partici- pants reported seeing flashes in the masks, but they were unable to identify any of the actual primes so they were not excluded from the following analyses. Results with or without these 4 participants did not differ.
Implicit Self-Stereotyping
If White Americans are subject to implicit self- stereotyping, they should respond to White stereotypes faster after self-relevant primes than neutral primes. In addition, they should respond equally fast, or slower, to African American stereotypes after self-relevant primes versus neutral primes. As in Experiment 1, to examine this prediction the transformed latencies were entered into a 2 (prime: self vs. neutral primes) � 2 (traits: African American vs. White American stereo- types) � 2 (valence: positive vs. negative words) repeated measure ANOVA. On average, participants responded faster to stereotypes of African Americans (M ¼ 677, SD ¼ 208) than stereotypes of White Americans (M ¼ 736, SD ¼ 233), F(1, 27) ¼ 12.93, p < .01, partial g2 ¼ .32. In addition, they responded faster to positive words (M ¼ 684, SD ¼ 198) than negative words (M ¼ 730, SD ¼ 239), F(1, 27) ¼ 15.68, p < .01, partial g2 ¼ .37. Each of these effects was moderated by prime.
As predicted, and consistent with Experiment 1, there was an interaction between prime and traits, F(1, 27) ¼
7.14, p ¼ .01, partial g2 ¼ .21 (see Figure 2). Simple effect analyses revealed that White participants responded faster to in-group stereotypes following self primes (M ¼ 705, SD ¼ 210) than to the same stereotypes following neutral primes (M ¼ 768, SD ¼ 276), t(27) ¼ 2.25, p ¼ .03, partial g2 ¼ .16. They also responded slower to African American stereotypes following self primes (M ¼ 695, SD ¼ 243) than to the same stereotypes follow- ing neutral primes (M ¼ 659, SD ¼ 190), though the differ- ence was only marginally reliable, t(27)¼ 1.91, p ¼ .07, partial g2 ¼ .12. This pattern was not moderated by valence, F(1, 27) < 1. In sum, White Americans showed implicit self-stereotyping on positive and negative in-group stereotypic traits, similar to the findings with women, and these associations were not affected by valence of the traits.
There was also an interaction between prime and valence, showing that participants were more likely to implicitly associate positive traits with their self- concepts than negative traits, F(1, 27) ¼ 8.42, p ¼ .01, partial g2 ¼ .24. Simple effect analyses showed that participants responded faster to positive traits following self primes (M ¼ 661, SD ¼ 201) than to the same traits following neutral primes (M ¼ 706, SD ¼ 216), t(27) ¼ 2.27, p ¼ .03, partial g2 ¼ .17. They also responded slower to negative traits following self primes (M ¼ 739, SD ¼ 248) than to the same negative traits following neutral primes (M ¼ 721, SD ¼ 245), t(27) ¼ 1.80, p ¼ .08, partial g2 ¼ .11.
Explicit self-stereotyping
As in Experiment 1, explicit self-ratings on stereoty- pic traits were submitted to a 2 (traits: African American vs. White American stereotypes) � 2 (valence: positive vs. negative words) repeated measure ANOVA. In gen- eral, participants rated themselves significantly higher on positive traits (M ¼ 4.68, SD ¼ 0.69) than negative
FIGURE 2 Average response latencies of White Americans to White
and Black stereotypes after self versus neutral primes.
IMPLICIT SELF-STEREOTYPING 123
traits (M ¼ 2.39, SD ¼ 0.64), F(1, 27) ¼ 218.80, p < .01, partial g2 ¼ .89. There was also a marginal interaction between these factors, F(1, 27) ¼ 2.87, p ¼ .10, partial g2 ¼ .10. Participants rated themselves equally high on positive stereotypes of White Americans (M ¼ 4.68, SD ¼ 0.91) and positive stereotypes of African Ameri- cans (M ¼ 4.68, SD ¼ 0.88), t(27) < 1. However, they rated themselves somewhat lower on negative stereo- types of African Americans (M ¼ 2.26, SD ¼ 0.71) than negative stereotypes of White Americans (M ¼ 2.53, SD ¼ 0.93) but the difference was not reliable, t(27) ¼ 1.36, p ¼ .19. These results suggest that White Americans are not significantly more likely to explicitly endorse positive in-group stereotypic traits than positive out-group stereotypic traits as characteristic of the self.
Positive and Negative Self-Stereotyping on Implicit Versus Explicit Measures
As in Experiment 1, we compared positive versus nega- tive self-stereotyping on the implicit and explicit measure by computing the Cohen’s ds in the same manner as in Experiment 1. Two difference scores were computed to indicate the degree to which participants explicitly self-stereotyped on positive and negative traits (self-ratings on positive White American stereotypes – self-ratings on positive African American stereotypes; self-ratings on negative White stereotype – self-ratings on negative African American stereotypes). Then, we derived the Cohen’s d from the mean difference and pooled standard deviation of these two difference scores. For the implicit measure, we first computed four scores that captured the degree to which participants implicitly associated positive and negative, White American and African American stereotypes to the self. These scores were created by subtracting latencies in response to a par- ticular type of stereotypes preceded by self primes from the latencies to the same set of traits preceded by neutral primes. With these four scores, we computed another two difference scores to indicate the degree to which partici- pants implicitly self-stereotyped positively and negatively (e.g., implicit self-stereotyping on positive White Ameri- can stereotypes – implicit self-stereotyping on positive African American stereotypes). Finally we computed the Cohen’s d based on the mean difference and the pooled standard deviation of these difference scores. A positive value of d indicates positive self-stereotyping and a negative value indicates negative self-stereotyping on both types of measures. On the explicit measure, par- ticipants were more likely to apply negative in-group traits to the self than positive in-group traits controlling for their self-ratings on outgroup stereotypes (d ¼ .60). This average d score was reliably different from 0, t(27) ¼ 2.42, p ¼ .02. Consistent with the previous analysis, the valence difference observed with explicit
self-ratings was substantially reduced when it was mea- sured implicitly (d ¼ .15); the average d score was not reli- ably different from 0, t(27) < 1. Furthermore, a paired-sample t test showed that White participants nega- tively self-stereotype to a greater extent on the explicit than implicit measure, t(27) ¼ 2.29, p ¼ .03.
Relationship between Explicit and Implicit Self-Stereotyping
As in Experiment 1, we correlated the four individual difference scores of implicit self-stereotyping on positive and negative, in-group and out-group traits computed for the previous analysis to the corresponding explicit self-ratings. In general, there was no reliable relationship between corresponding implicit and explicit measures on any type of traits. Correlation coefficients for positive stereotypes and negative stereotypes of White Americans, and positive and negative stereotypes of African Americans were r(28) ¼ .19, .12, .09, and .07, respectively, all ps > .30. Therefore, as in Experiment 1, there was no consistent correspondence between implicit and explicit self-stereotyping.
Implicit Self-Stereotyping and Self-Esteem
Finally, we explored whether the four individual dif- ference scores representing implicit association between in-group and out-group stereotypic traits and the self were related to self-esteem. We first correlated the degree to which participants associated positive and negative in-group stereotypic traits with the self to global and state self-esteem. There was no reliable relationship between negative implicit self-stereotyping and global or state self-esteem (r ¼ .06 and r ¼ .07, ps > .70, respectively). But positive implicit self-stereotyping seemed to relate to self-esteem. Implicit self-stereotyping on positive traits was associated with higher state self-esteem (r ¼ .42, p ¼ .03). The relationship with global self-esteem was in the same direction but weaker (r ¼ .30, p ¼ .12). There- fore, positive but not negative implicit self-stereotyping showed relationships with one’s self-esteem. In contrast, we found no reliable relationship between the degrees to which White participants associated out-group traits with the self and global or state self-esteem (all ps > .40). These exploratory findings suggest that implicit self-stereotyping, as opposed to associations between out-group traits and the self, can manifest in other self-relevant judgments, but this relationship may be limited to attributes with positive implications.
GENERAL DISCUSSION
In two experiments, we showed that women and White Americans implicitly self-stereotype on both positive
124 LUN, SINCLAIR, COGBURN
and negative traits. Although it might not be surprising that members of a social group would implicitly associ- ate positive in-group stereotypic traits with the self-concept, the fact that they also implicitly self-stereotype on negative traits is novel and consistent with the notion that people implicitly associate in-group traits with the self regardless of the valence of the traits (e.g., Jost, 2001; Smith & Henry, 1996; Turner et al., 1987). Furthermore, it is notable that White Americans, a group that is not typically perceived as a target of stereotypes, were found to engage in implicit self- stereotyping at all. These findings illustrate that perva- sive cultural stereotypes subtly shape the self-concept without individuals’ conscious awareness. Moreover, unlike much of the previous work that focused on indi- vidual differences (Greenwald et al., 2002; but see Greenwald & Farnham, 2000), our research showed that implicit self-stereotyping also occurs at a group level.
In contrast, evidence of explicit self-stereotyping on these traits painted a less consistent picture. In Experi- ment 1, women rated themselves higher on positive feminine traits than other types of traits. In Experi- ment 2, White Americans rated themselves equally high on positive in-group and out-group stereotypic traits but they were also least likely to endorse out-group negative stereotypic traits as self-descriptive. Although women and White Americans did not show a reliable difference between implicit self-stereotyping on positive versus negative traits, valence difference emerged on the expli- cit measure. Hence, members of a social group may be able to consciously eschew negative in-group stereotypes as characteristics of the self, but they may still noncon- sciously associate these traits with the self.
It is important to note that implicit and explicit self-stereotyping was tested with one particular social group in each experiment. The lack of a comparison group (i.e., men, African Americans) invites an alterna- tive interpretation about why there was no valence effect on implicit self-stereotyping. Such interpretation, for example, would suggest that women implicitly self-stereotyped on both positive and negative traits because feminine traits were simply more applicable to the self-concept than masculine traits. We believe that our results on explicit self-stereotyping cast some doubts to this interpretation. If these traits were simply more applicable, we would expect to find parallel results on the explicit measure, but that was not the case. Although this contention does not completely rule out the alter- native interpretation, it is less likely that such interpreta- tion would explain the same finding with two different social groups and two different sets of stereotypes.
Not only did implicit and explicit self-stereotyping yield different results, but there was also a general lack of correlation between these measures for women and White Americans. This suggests that individual
differences in people’s conscious self-perceptions on in-group attributes do not have a direct relationship with how much they implicitly associate those attributes with the self. This inconsistency suggests that implicit and explicit measures may capture different aspects of one’s self-concept (Bosson, Swann, & Pennebaker, 2000; Spalding & Hardin, 1999). In light of recent per- spectives on the correspondence between implicit and explicit measures (Nosek, 2007), future research should examine whether there are moderating factors that influence the relationship between implicit and explicit self-concepts such as the centrality of traits to the self-concepts, identification with the in-group (Chen, Chen, & Shaw, 2004) or different cultural beliefs about self-perceptions (Hetts, Sakuma, & Pelham, 1999).
Although we did not find consistent correspondence between explicit and implicit self-stereotyping, we found a moderate relationship between positive, but not nega- tive, implicit self-stereotyping and self-esteem among White Americans. This finding suggests that even though implicit self-stereotyping occurs on positive and negative traits, it does not have to be costly to one’s self-esteem. People may still preserve their positive self-regard by dissociating negative in-group stereotypes from their sense of self-worth. Previous research has demonstrated a similar self-serving cognitive balancing among members of socially devalued or low-status groups, but they do so through suppressing positive evaluations toward the in-group and thus by showing out-group favoritism (Lane, Mitchell, & Banaji, 2005; Rudman et al., 2002). The out-group favoritism was thought to result from a propensity to reduce cognitive inconsistency engendered by positive self-regards and the negative implications of being the members of a low-status group (Lane et al., 2005). Although past research focuses on low-status groups and implicit group evaluation, the current research suggests that high-status group members also engage in a similar cognitive balancing act because positive and negative in-group stereotypes are being incorporated into the self-concepts, as the present research has shown. Never- theless, the current finding regarding the relationship between implicit self-stereotyping and self-esteem is preliminary. Its correlational nature leaves open the interpretation that people who have higher self-esteem tend to associate positive in-group traits with the self.
Future research should examine the relationship between implicit self-stereotyping and other self-relevant measures in greater detail. For example, one may expect the relationship between implicit self-stereotyping and self-esteem to be moderated by the perceived importance of the traits in question. Past research suggests that the importance individuals place on a specific self-view determines the degree to which the self-view influences explicit self-esteem (Pelham, 1995; Pelham & Swann,
IMPLICIT SELF-STEREOTYPING 125
1989). If people are deemphasizing the centrality of certain stereotypes to protect self-esteem (Crocker & Major, 1989; Major & Schmader, 1998), we would expect perceived centrality of in-group stereotypes to be related to the degree to which an individual implicitly associates these traits with the self.
In addition to self-evaluations, future work on implicit self-stereotyping may also help shed light on the mechan- isms through which targets behave in a stereotype- consistent manner (see Wheeler & Petty, 2001, for a review). Research has demonstrated that targets are more likely to behave in a stereotype-consistent manner when in-group stereotypes are cognitively accessible (e.g., Levy, 1996; Steele & Aronson, 1995) or when these stereotypes are made relevant to the self (Marx, Stapel, & Muller, 2005; Wheeler, Demarree, & Petty, 2005; Wheeler, Jarvis, & Petty, 2001). Together, these findings suggest that the degree to which individuals implicitly self-stereotype may moderate the extent to which a target will behave in a stereotype-consistent manner in a given situation.
In sum, consistent with the notion of influential the- ories like the looking-glass self (e.g., Cooley, 1902) and recent research on implicit cognition (e.g., Greenwald & Banaji, 1995), cultural stereotypes of groups to which an individual belongs, including those with negative connotations, are incorporated into the self-concept. Moreover, these associations between the self and ste- reotypes are not subject to conscious introspection. As a society continuously reinforces stereotypes of different social groups, these beliefs may subtly shape the self-concept of members of these groups without them being consciously aware of it.
REFERENCES
Allport, G. W. (1979). The nature of prejudice. New York:
Addison-Wesley. (Original work published 1954).
Bargh, J. A., & Chartrand, T. L. (2000). The mind in the middle:
A practical guide to priming and automaticity research. In H. T.
Reis & C. M. Judd (Eds.), Handbook of research methods in social
and personality psychology (pp. 253–285). New York: Cambridge
University Press.
Bargh, J. A., & Pietromonaco, P. (1982). Automatic information pro-
cessing and social perception: The influence of trait information pre-
sented outside of conscious awareness on impression formation.
Journal of Personality and Social Psychology, 43, 437–449.
Bosson, J. K., Swann, W. B. Jr., & Pennebaker, J. W. (2000). Stalking
the perfect measure of implicit self-esteem: The blind men and the
elephant revisited? Journal of Personality and Social Psychology,
79, 631–643.
Chen, S., Chen, K. Y., & Shaw, L. (2004). Self-verification motives at
the collective level of self-definition. Journal of Personality and
Social Psychology, 86, 77–94.
Cooley, C. H. (1902). In human nature and the social order. New York:
Schocken Books.
Crocker, J., & Major, B. (1989). Social stigma and self-esteem: The
self-protective properties of stigma. Psychological Review, 96, 608–630.
Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. T. Gilbert,
S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology
(Vol. 2, 4th ed., pp. 504–553). New York: McGraw-Hill.
Darley, J. M., & Gross, P. H. (1983). A hypothesis-confirmation bias
in labeling effects. Journal of Personality and Social Psychology, 44,
20–33.
Dasgupta, N. (2004). Implicit ingroup favoritism, outgroup favoritism, and
their behavioral manifestations. Social Justice Research, 17, 143–169.
Dasgupta, N., McGhee, D. E., Greenwald, A. G., & Banaji, M. R.
(2000). Automatic preference for White Americans: Eliminating
the familiarity explanation. Journal of Experimental Social
Psychology, 36, 316–328.
Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and
controlled components. Journal of Personality and Social Psychol-
ogy, 56, 5–18.
Devine, P., & Elliot, A. (1995). Are racial stereotypes really fading?
The Princeton trilogy revisited. Personality and Social Psychology
Bulletin, 21, 1139–1150.
Devos, T., & Banaji, M. R. (2003). Implicit self and identity. In M. R.
Leary & J. P. Tangney (Eds.), Handbook of self and identity (pp. 91–
105). New York: Guilford.
Dunham, Y., Baron, A. S., & Banaji, M. R. (2006). From American
city to Japanese village: A cross-cultural investigation of implicit
race attitudes. Child Development, 77, 1268–1281.
Fazio, R. (1990). A practical guide to the use of response latency in
social psychological research. In C. Hendrick & M. S. Clark
(Eds.), Review of personality and social psychology: Vol. 11. Research
methods in personality and social psychology (pp. 74–97). Newbury
Park, CA: Sage.
Fiske, S. T. (1998). Stereotyping, prejudice, and discrimination. In D.
T. Gilbert & S. T. Fiske (Eds.), The handbook of social psychology
4th ed., Vol. 2 (pp. 357–411). Boston: McGraw-Hill.
Fiske, S. T., Cuddy, A., Glick, P., & Xu, J. (2002). A model of (often
mixed) stereotype content: Competence and warmth respectively
follow from perceived status and competition. Journal of Personality
and Social Psychology, 82, 878–902.
Fiske, S. T., Lin, M., & Neuberg, S. L. (1999). The continuum model:
Ten years later. In S. Chaiken & Y. Trope (Eds.), Dual-process the-
ories in social psychology (pp. 231–254). New York: Guilford.
Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: Atti-
tudes, selfesteem and stereotypes. Psychological Review, 102, 4–27.
Greenwald, A. G., Banaji, M. R., Rudman, L. A., Farnham, S. D.,
Nosek, B. A., & Mellott, D. S. (2002). A unified theory of implicit
attitudes, stereotype, selfesteem, and self-concept. Psychological
Review, 109, 3–25.
Greenwald, A. G., & Farnham, S. D. (2000). Using the implicit
association test to measure self-esteem and self-concept. Journal of
Personality and Social Psychology, 79, 1022–1038.
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998).
Measuring individual differences in implicit cognition: The implicit
association test. Journal of Personality & Social Psychology, 74,
1464–1480.
Hamilton, D., Sherman, J., & Ruvolo, C. M. (1990).
Stereotyped-based expectancies: Effects on information processing
and social behavior. Journal of Social Issues, 46, 36–60.
Heatherton, T. F., & Polivy, J. (1991). Development and validation of
a scale for measuring state selfesteem. Journal of Personality and
Social Psychology, 60, 895–910.
Hetts, J. J., Sakuma, M., & Pelham, B. W. (1999). Two roads to posi-
tive regard: Implicit and explicit selfevaluation and culture. Journal
of Experimental Social Psychology, 35, 512–559.
Hilton, J. L., & von Hippel, W. (1996). Stereotypes. Annual Review of
Psychology, 4, 237–271.
Jost, J. T. (2001). Outgroup favoritism and the theory of system justi-
fication: A paradigm for investigating the effects of socioeconomic
126 LUN, SINCLAIR, COGBURN
success on stereotype content. In G. B. Moskowitz (Ed.), Cognitive
social psychology: The Princeton symposium on the legacy and future
of social cognition (pp. 89–102). Mahwah, NJ: Erlbaum.
Jost, J. T., & Banaji, M. R. (1994). The role of stereotyping in
system-justification and the production of false consciousness.
British Journal of Social Psychology, 33, 1–27.
Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004). A decade of system
justification theory: Accumulated evidence of conscious and uncon-
scious bolstering of the status quo. Political Psychology, 25, 881–920.
Jost, J. T., & Burgess, D. (2000). Attitudinal ambivalence and the con-
flict between group and system justification motives in low status
groups. Personality and Social Psychology Bulletin, 26, 293–305.
Jost, J. T., Pelham, B. W., & Carvallo, M. R. (2002). Nonconscious
forms of system justification: Implicit and behavioral preferences
for higher status groups. Journal of Experimental Social Psychology,
38, 586–602.
Katz, D., & Braly, K. (1933). Racial stereotypes of one hundred
college students. Journal of Abnormal Psychology, 3, 175–193.
Kay, A. C., & Jost, J. T. (2003). Complementary justice: Effects of
‘poor but happy’ and ‘poor but honest’ stereotype exemplars on
system justification and implicit activation of the justice motive.
Journal of Personality and Social Psychology, 85, 823–837.
Lane,K.A.,Mitchell,J.P.,&Banaji,M.R.(2005).Meandmygroup:Cultural
status can disrupt cognitive consistency. Social Cognition, 23, 353–386.
Levy, B. (1996). Improving memory in old age through implicit
self-stereotyping. Journal of Personality & Social Psychology, 71,
1092–1107.
Macrae, C. N., & Bodenhausen, G. V. (2000). Social cognition:
Thinking categorically about others. Annual Review of Psychology,
51, 93–120.
Major, B., & Schmader, T. (1998). Coping with stigma through
psychological disengagement. In J. K. Swim & C. Stangor (Eds.),
Prejudice: The target’s perspective (pp. 219–241). San Diego, CA:
Academic Press.
Marx, D. M., Stapel, D. A., & Muller, D. (2005). We can do it: The
interplay of construal orientation and social comparisons under
threat. Journal of Personality and Social Psychology, 88, 432–446.
Mead, G. H. (1934). Mind, self and society. Chicago: University of
Chicago Press.
Mullen, B., Brown, R., & Smith, C. (1992). Ingroup bias as a function
of salience, relevance, and status: An integration. European Journal
of Social Psychology, 22, 103–122.
Nosek, B. A. (2007). Implicit-explicit relations. Current Directions in
Psychological Science, 16, 65–69.
Nosek, B. A., & Banaji, M. R. (2001). The go=no-go association task.
Social Cognition, 19, 625–664.
Nosek, B. A., Banaji, M. R., & Greenwald, A. G. (2002a). Harvesting
implicit group attitudes and beliefs from a demonstration web site.
Group Dynamics: Theory, Research, & Practice, 6, 101–115.
Nosek, B. A., Banaji, M. R., & Greenwald, A. G. (2002b).
Math ¼ male, me ¼ female, therefore math ¼ = ¼ me. Journal of Personality and Social Psychology, 83, 44–59.
Olivier, K., & Snyder, M. (2003). Stereotypes and behavioral confirma-
tion: From interpersonal to intergroup perspectives. In M. Zanna
(Ed.), Advances in Experimental Social Psychology (Vol. 35,
pp. 153–234). New York: Academic Press.
Pelham, B. W. (1995). Self-investment and self-esteem: Evidence for a
Jamesian model of selfworth. Journal of Personality & Social
Psychology, 69, 1141–1150.
Pelham, B. W., & Swann, W. B. (1989). From self-conceptions to
self-worth: On the sources and structure of global self-esteem. Jour-
nal of Personality & Social Psychology, 57, 672–680.
Perdue, C. W., Dovidio, J. F., Gurtman, M. B., & Tyler, R. B. (1990).
Us and them: Social categorization and the process of intergroup
bias. Journal of Personality and Social Psychology, 59, 475–486.
Ratcliff, R. (1993). Methods for dealing with reaction time outliers.
Psychological Bulletin, 114, 510–532.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
NJ: Princeton University Press.
Rudman, L. A. (2004). Sources of implicit attitudes. Current Directions
in Psychological Science, 13, 79–82.
Rudman, L. A., Feinberg, J., & Fairchild, K. (2002). Minority mem-
bers’ implicit attitudes: Automatic ingroup bias as a function of
group status. Social Cognition, 20, 294–320.
Rudman, L. A., Greenwald, A. G., & McGhee, D. E. (2001). Implicit
self-concept and evaluative implicit gender stereotypes: Self and
in-group share desirable traits. Personality and Social Psychology
Bulletin, 27, 1164–1178.
Simon, B., Glassner-Bayerl, B., & Stratenwerth, I. (1991). Stereotyping
and self-stereotyping in a natural intergroup context: The case of
heterosexual and homosexual men. Social Psychology Quarterly,
54, 252–266.
Simon, B., & Hamilton, D. (1994). Self-stereotyping and social
context: The effects of relative in-group size and in-group status.
Journal of Personality and Social Psychology, 66, 699–711.
Smith, E. R., Coats, S., & Walling, D. (1999). Overlapping mental
representations of self, in-group, and partner: Further response time
evidence and a connectionist model. Personality and Social Psychol-
ogy Bulletin, 25, 873–882.
Smith, E. R., & Henry, S. (1996). An in-group becomes part of the self:
Response time evidence. Personality and Social Psychology Bulletin,
22, 635–642.
Spalding, L. R., & Hardin, C. D. (1999). Unconscious unease and
self-handicapping: Behavioral consequences of individual differ-
ences in implicit and explicit self-esteem. Psychological Science, 10,
535–539.
Srull, T. K., & Wyer, R. S. (1979). The role of category accessibility in
the interpretation of information about persons: Some determinants
and implications. Journal of Personality & Social Psychology, 37,
1660–1672.
Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellec-
tual test performance of African Americans. Journal of Personality
& Social Psychology, 69, 797–811.
Tajfel, H., & Turner, J. (1986). The social identity theory of inter-
group behavior. In S. Worchel & W. Austin (Eds.), Psychology
of intergroup relations (pp. 59–86). Monterey, CA: Brooks=Cole.
Tice, D. M., & Wallace, H. M. (2003). The reflected self: Creating
yourself as (you think) others see you. In M. R. Leary & J. P.
Tangney (Eds.), Handbook of self and identity (pp. 91–105). New
York: Guilford.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell,
M. S. (1987). Rediscovering the social group: A self-categorization
theory. Oxford, UK: Basil Blackwell.
Wheeler, S. C., Demarree, G. K., & Petty, R. E. (2005). The roles of the
self in primingtobehavior effects. In A. Tesser, J. V. Wood, & D. A.
Stapel (Eds.), On building, defending and regulating the self: A psycho-
logical perspective (pp. 245–271). New York: Psychology Press.
Wheeler, S. C., Jarvis, W. B. G., & Petty, R. E. (2001). Think unto
others: The selfdestructive impact of negative racial stereotypes.
Journal of Experimental Social Psychology, 37, 173–180.
Wheeler, S. C., & Petty, R. E. (2001). The effect of stereotype activa-
tion on behavior: A review of possible mechanisms. Psychological
Bulletin, 127, 797–826.
Wittenbrink, B., Judd, C. M., & Park, B. (1997). Evidence for
racial prejudice at the implicit level and its relationship with ques-
tionnaire measures. Journal of Personality & Social Psychology,
72, 262–274.
Word, C. O., Zanna, M. P., & Cooper, J. (1974). The nonverbal media-
tion of selffulfilling prophecies in interracial interaction. Journal of
Experimental Social Psychology, 10, 109–120.
IMPLICIT SELF-STEREOTYPING 127