xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001461874
007 cr mnu|||uuuuu
008 040406s2004 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000231
Hypocrisy induction to alter selection decisions among aversive racists
h [electronic resource] :
analyzing the role of external motivation to respond without prejudice /
by Andrew Biga.
[Tampa, Fla.] :
University of South Florida,
Thesis (M.A.)--University of South Florida, 2004.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 76 pages.
ABSTRACT: The present study will examine the effects of hypocrisy induction on selection decisions for two populations: Aversive Racists and truly non-prejudiced individuals. Aversive Racists are operationally defined as individuals who are low in explicit prejudice and high in implicit prejudice, whereas truly non-prejudiced individuals are defined as being low in both explicit and implicit prejudice. These two groups of people will differ on their ratings of job applicants, so that Aversive Racists will rate Black applicants lower than White applicants (with comparable job credentials) while truly non-prejudiced individuals will rate them similarly. The induction of hypocrisy will serve as a manipulation that reverses Aversive Racists ratings of job applicants; Black applicants will now be rated higher than White applicants with similar job credentials. External motivation to respond without prejudice will moderate these effects in the expected direction.
Adviser: Vandello, Joseph
t USF Electronic Theses and Dissertations.
Hypocrisy Induction to Alter Selection Decisions among Aver sive Racists: Analyzing the Role of External Mo tivation to Respond without Prejudice by Andrew Biga A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Joseph Vandello, Ph.D. Tammy Allen, Ph.D. Toru Shimizu, Ph.D. Date of Approval: March 2, 2004 Keywords: selection, prejudice, implicit, explicit, motivation Copyright 2004, Andrew Biga
i Table of Contents List of Tables iii List of Figures iv Abstract v Introduction 1 Historical Perspectives on the Study of Prejudice 2 Aversive Racism 4 Implicit and Explicit Attitudes 5 Implicit and Explicit Attitudes within the Aversive Racism Framework 8 Internal and External Motivation 10 Hypocrisy Induction as a Mech anism to Reduce Prejudice 12 Present Study 14 Hypotheses 15 Method 21 Participants 21 Materials 22 Explicit prejudice 22 Implicit prejudice 22 Internal and External Motivation 24 Job applicant ratings 25 Design 26 Procedure 27 Analyses 30 Results 31 Hypotheses 31 Hypothesis 1 31 Hypothesis 2 32 Hypothesis 3 33 Hypothesis 4 33 Hypothesis 5 34 Additional Analyses 34
ii Hypocrisy Effect: Mediation by Emotions 34 Individual Differences 36 Discussion 38 References 44 Appendices 50 Appendix A: Modern Racism Scale 51 Appendix B: Internal/External Motiva tion to Respond without Prejudice 52 Appendix C: Business Hiring Decision 53 Appendix D: Hypocrisy Manipulation 62 Appendix E: Positive Affect / Negative Affect Scales 65
iii List of Tables Table 1 Correlation matrix fo r independent variables 66 Table 2 Correlation matrix for dependent variables 67 Table 3 Cell means for (hypocrisy x applicant race) interaction 68
iv List of Figures Figure 1. Hypothesis 1 16 Figure 2. Hypothesis 2 17 Figure 3. Hypothesis 3 18 Figure 4. Hypothesis 4 19 Figure 5. Hypothesis 5 20 Figure 6. Distribution of IAT Scores 68 Figure 7. Mediation of hypocrisy 69
v Hypocrisy Induction to Alter Selecti on Decisions among Av ersive Racists: Analyzing the Role of External Mo tivation to Respond without Prejudice Andrew Biga ABSTRACT The present study will examine the effects of hypocrisy i nduction on selection decisions for two populations: Aversive Racists and tr uly non-prejudiced indi viduals. Aversive Racists are operationally defi ned as individuals who are low in explicit prejudice and high in implicit prejudice, whereas truly nonprejudiced individuals are defined as being low in both explicit and implicit prejudice. These two groups of people will differ on their ratings of job applicants so that Aversive Racists wi ll rate Black applicants lower than White applicants (with comparable job credentials) while truly non-prejudiced individuals will rate them similarly. The induction of hypocrisy will serve as a manipulation that reverses Aver sive Racists ratings of job applicants; Black applicants will now be rated higher than White applicants with similar job credentials. External motivation to respond without pr ejudice will moderate thes e effects in the expected direction.
1 Introduction The expression of racism has changed dramatically over the last century. Social norms now dictate that overt forms of racial prej udice are unacceptable, while more covert forms have emerged (Dovidi o, 2001). Racial attitudes among Whites have generally become more liberal in the past half-century, and it is now the norm to support broad principles of equality (Sch uman & Krysan, 1999; Crandall, Eshleman, & OÂ’Brien, 2002). As egalitarian belie fs among Whites have become more prominent, obvious discrimination against Blacks has become unacceptable by todayÂ’s legal and social standards. This has ha d a large impact upon hiring procedures used in organizati ons, but there still may be sub tle biases that affect these processes. Selection issues are a core area of re search in Industrial/Organizational Psychology. Racial prejudice may still bias how employers and organizations select employees (Dovidio & Gaertner, 2000). Pr ejudice can occur w ithout an individual being aware of its influence. This basic premise has led to research on more subtle forms of prejudice that have emerged under the existing social norms that prohibit overt expressions of r acial intolerance. Covert forms of racial prejudice are no t as easily detected, but still have a major impact on our society (Gaertner & Dovidio, 2000). Traditional self-report
2 measures of racial prejud ice do not adequately quantify subtle forms of racial prejudice. Racism has changed conceptua lly and new measures have been developed and researched that tap into this change, su ch as measures of im plicit racial attitudes and motivation to respond without prejudi ce. The following introduction will cover four topics. First, the concept of Aver sive Racism (Dovidio & Gaertner, 1986) will serve as a framework to assist in understan ding how subtle prejudice manifests itself in selection decisions. Second, the distinct ion between implicit and explicit attitudes will be discussed in the context of the Aversive Racism model. Third, different motivational aspects of prejudice will be examined, focusing specifically on the motivation to respond without prejudice and how this factor may moderate selection decisions. Fourth, the induction of hypocrisy will be reviewed as a means to change prejudiced behavior. The goal of this rese arch is to determin e how the induction of hypocrisy is related to indi vidual characteristics, such as implicit and explicit prejudice, in the rating of job applicants for a high status job position. External motivation to respond without prejudice will be examined to determine if this individual variable may moderate the in teraction between hypoc risy induction and Aversive Racism. Historical Perspectives on the Study of Prejudice In order to understand the contemporary view towards racial prejudice, it is important to look back at past research on the subject. Dovidio (2001) identifies three waves of research that help to explain th e current trend in the analysis of racial prejudice. The first wave, from the 1920Â’s through 1950Â’s, viewed racial prejudice as
3 a deviant pathology. With roots in Freudian theory, the authoritarian personality was seen as a major indicator of racial prejudice (Adorno, Frenke l-Brunswik, Levinson, & Sanford, 1950). Authoritarian personality places a large emphasis on order and respect, which manifests itself in high obedi ence to authority figures. Authority is not to be questioned, and these individuals valu e strict rules and regulations. In part because prejudice was far more widespr ead and normative, the authoritarian personality proved inadequate to explain prejudice. Coinciding with the emerging cognitive revolution in social psychology in the 1970Â’s and in contrast to the first wave, th e second wave of research viewed racial prejudice as a normative process and the resu lt of an adaptive cognitive process in which a person attempts to simplify an im mensely complex world. Stereotypes serve a function that is normal and necessary. W ithout the ability to quickly classify people into meaningful categories, individuals would have an extremely difficult time interacting without us ing this simplification tool. Several similar theoretical perspectives emerged during the mid-1980s to describe prejudice as more subtle and normative. Theories of aversive racism (Dovidio & Gaertner, 1986), modern racism (McConahay, 1986), and symbolic racism (S ears, 1988) were developed to explain the changing trend in racial attitudes that overt measures of racial prejudice were reflecting. These theories all share the theme that Whites experience a conflict between non-prejudiced values and prejudi ced tendencies. Gaertner and Dovidio (1986) provided significant evidence that discrimination was stil l occurring in more subtle ways that would not have been dete cted by overt measures of racial prejudice
4 by using experiments that focused on helpi ng behavior, judicial court decisions, and reaction times to positive and negative words paired with Â“BlacksÂ” and Â“Whites.Â” In experiments like these, racial prejudice s eemed to have become more covert and subtle in nature, while significantly affec ting the behavior of individuals (Dovidio & Gaertner, 1986; McConahay, 1986). Aversive Racism Dovidio and Gaertner (1986) introdu ced the idea of Aversive Racism to explain the contemporary nature of discri mination and prejudicial attitudes that categorize many White Americans. Overt Â“old-fashionedÂ” racism is less accepted today than it once was in American society. A social norm against the expression of prejudicial beliefs and discriminatory actions against Black Americans has developed. The construct of Aversive Racism was originally proposed to describe White Americans that have strong egalitarian id eals and support non-prej udicial values, but who tend to find interactions with Blacks unpl easant and who hold some level of antiBlack sentiment. These individuals do not consider themselves to be prejudiced, but research shows that a pattern of discrimi natory behavior is evident in certain situations (Gaertner & Dovidio, 2000). Aver sive racism can only be detected when the social norms that dictate how to interact are unclear. Ambiguity allows aversive racists to act in a discriminatory fashion b ecause they can attribut e their decisions to alternative, non-prejudic ed reasons. Aversive racists are able to justify their behavior by rationalizing that their actio ns are appropriate. While the construct of Aversive Racism has been applied to other minor ity groups (Dovidio, Gaertner, Anastasio, &
5 Sanitioso, 1992; Son Hing, Li, & Zanna, 2002), the majority of research on Aversive Racism has focused on White attitudes towa rd Blacks, and will be the focus of the proposed study. Measuring racial prejudice became more difficult with the development of the social norms against its expression. A need existed for improved measures to address this issue. More covert measures, such as the Modern Racism Scale (McConahay, 1986), were developed to deal with the shift in the nature of racial prejudice. This scale is intended to be non-reactive, so th at a more accurate description of a personÂ’s attitudes could be measured in contrast to earlier scales. N on-reactive scales are designed to avoid social desi rability bias by wording th e questions so as to not directly reveal the cont ent of the measure (Fazio, Jackson, Dunton, & Williams, 1995). The Modern Racism Scale takes into account the change in norms regarding the expression of racial prejudice; whet her it promotes non-reactivity in respondents has recently been debated (Fazio, Jackson, Dunton, & Williams, 1995). In order to avoid the issu e of reactivity, recent measures have been created that attempt to circumvent conscious pr ocessing by using reaction times to measure the strength of associations between the categories of Â“WhiteÂ” and Â“BlackÂ” with positive or negative words. These types of measures have sparked the third wave of research on racial prejudice. Implicit and Explicit Attitudes The third wave of research began re cently in the 1990s, with a focus on the multidimensional features of racial prej udice (Dovidio, 2001). The concepts of
6 implicit and explicit attitudes were incor porated into the theoretical framework established in the second wave. Implicit attitudes are defined to be Â“introspectively unidentified (or inaccurately identified) traces of past experience that mediate favorable or unfavorable feeling, thought, or action toward social objectsÂ” (Greenwald & Banaji, 1995, p. 8). Theoreti cally, implicit attitudes are automatic and outside the awareness of the individual. In contrast, explicit attitudes are conscious and under the control of the individual. One such model that explains the co-existence of two attitudes is the Model of Dual Attitudes, proposed by Wilson, Li ndsey, and Schooler (2000). The model postulates that a personÂ’s at titudes can have two dimensions, one implicit and one explicit. These two attitudinal dimensions do not have to agree and can coexist, such that possible disagreement does not necessarily cause a state of conflict (Wilson, Lindsey, & Schooler, 2000). An important distinction between dual attitudes and ambivalence is made by Wilson et al. (2000): where ambivalence can create a state of conflict, dual attitudes allow for the individua l to express the most accessible attitude. Implicit attitudes are thought to be automatic and difficult to control, while explicit attitudes require cognit ive effort to express. Becaus e explicit attitudes are under the conscious control of an individual, they can be changed fairly easily, while implicit attitudes are much more diffi cult to alter. This has implications for how attitudes affect behavior. Even when a person has the cognitive capacity to express an explicit attitude, implicit attitudes may unconsciously influen ce nonverbal behavior (Wilson, Lindsey, & Schooler, 2000).
7 Evidence has shown that e xplicit racial attitudes are related to self-perceptions of behavior in interracial in teractions, whereas implicit attitudes are related to actual nonverbal behavior, such as eye contact and blinking (Dovidio, Kawakami, Johnson, Johnson, & Howard, 1997; Dovidio, Ka wakami, & Gaertner, 2002). White individualsÂ’ perceptions of their own behavior were re lated to their explicit attitudes toward Blacks in an interra cial interaction. But, Black individualsÂ’ ratings of the White individualÂ’s behavior during the interaction was directly related to the White individualsÂ’ levels of im plicit racial prejudice (Dovi dio, Kawakami, & Gaertner, 2002). This evidence supports the conceptualizati on of implicit and explicit attitudes controlling different types of behavior. The distinction between implicit and e xplicit has fairly strong empirical support. A relatively new measure, the Implicit Association Test (IAT; Greenwald, McGhee, & Schwartz 1998), has been extensiv ely used to measure implicit attitudes. The basic idea behind the IAT is that if two concepts are strongly associated (e.g. Â“BlackÂ” with negative words), then there wi ll be a shorter response time than when the two concepts are incongruent with the stereotype (e.g. Â“BlackÂ” with positive words). Pressing keys on a keyboard are used to measure response times during a computerized administration of the IAT (G reenwald & Nosek, 2001). Research using this measure has explored the relationship between explicit and implicit measures. Explicit measures (e.g. Modern Racism Scale, feeling thermometer, semantic differential measures) demonstrate an average correlation of .25 with the IAT (Greenwald, McGhee, & Schwartz, 1998). Nosek, Banaji, and Greenwald (2002)
8 found the correlation between explicit and implicit measures ranges from.08 to.47 depending on which attitudes the IAT was measuring. Although the validity of the IAT has been questioned by some authors (Bosson, Swann, & Pennebaker, 2000; Lowery, Hardin, & Sinclair, 2001), its ability to measure an aspect of an individual that is distinct from explicit attitudes appears well established. The IAT, and similar measures, have shown evidence for the exis tence of implicit attitudes, Â“Thus weak correlations between explicit and implicit at titudes may not reflect weak measures, but may instead represent the nature of contemporary prejudiceÂ” (Dovidio, Kawakami, & Beach, 2001, p. 183). While imp licit attitudes are outside of conscious awareness, the contextual environment ha s been shown to have an effect on the expression of these attitudes (Lowery, Ha rdin, & Sinclair, 2001; Wittenbrink, Judd, & Park, 2001). Implicit and Explicit Attitudes within the Aversive Racism Framework Because traditional paper-and-pencil measures ask people to consciously reflect on their racial attit udes, these measures assess expl icit prejudice. With a new focus on implicit attitudes, the framework of Aversive Racism has been modified to more accurately reflect the cont emporary view of prejudice. Aversive Racists are currently operati onally defined as individuals who are low on explicit prejudice, but high on implicit prejudice (Son Hing, Li, & Zanna, 2002). Theoretically, high and low levels of implicit and exp licit prejudice can combine to form four cate gories of individuals as s hown in the following chart:
9 High Explicit Prejudice Low Explicit Prejudice High Implicit Prejudice Overt Racists Aversive Racists Low Implicit Prejudice Compliant Racists (small group) Truly Non-Prejudiced Compliant racists can theoretically exist, although this group is not typically very large in a sample. This group of individuals may interpret the social norms to support prejudice while holding conflicting internal at titudes. Historica lly, compliant racists may have been more abundant during the Jim Crow era when social norms established racial prejudice as the expected position of White Americans (Pettigrew, 1959). The present research is primarily con cerned with the two groups on the right of the chart, Aversive Racists and trul y non-prejudiced individuals. The reasoning behind this interest is that Aversive Racist s, who consciously reje ct prejudice, may be more able to alter their subtle biases if they are more aware of how these processes operate, and thus may be an opportune group to target for intervention. Reanalysis of the construct of Aver sive Racism has led to different interpretations of the original theory. In stead of anti-Black se ntiment being at the root of Aversive Racism, pro-White atti tudes might instead drive this process (Gaertner & Dovidio, 2000; Wittenbrink, Judd, & Park, 1997). Just because an action favors a personÂ’s in-group does not necessari ly indicate negative feelings towards a personÂ’s out-group. Gaertner and Dovidio (2000) investigat ed past research that
10 supports this conclusion. A previous study surveying Whites found that positive characteristics were associated more str ongly with Whites than with Blacks, whereas negative characteristics were not more asso ciated with Blacks than with Whites. Whites and Blacks were both rated simila rly on negative attr ibutes, but not on positive attributes (Gaertner & Dovidio, 2000). It seems that Whites may view other Whites in a positive manner, but view Blacks in a neutral rather than negative manner. In addition, Wittenbrink, Judd, and Park (1997) found that implicit measures of racial attitudes correlated with explicit measures of r acial attitudes except for the Anti-Black scale. Aversive racists are by definition low on explicit prejudice and high on implicit prejudice, so the findings that implicit racial attitudes did not correlate with the Anti-Black scale seem to support the reanalysis of Aversive Racism as being rooted in pro-White, not anti-Black, sentiment. This has implications for the prevention of racial discrimination where so cial norms that dictate the inhibition of negative actions toward Blacks may not be e ffective because the nature of Aversive Racism is pro-White and not anti-Black (Dovidio & Gaertner, 1998). The reanalysis of Aversive Racism needs to be incorporated into current research on racial prejudice that focuses on the ideas of implicit and explicit prejudice. Internal and External Motivation The strength of the social norm agai nst overt expression of prejudice makes measuring prejudice a difficult task; however, looking at prejudice from a motivational perspective has relevance. Plan t and Devine (1998) postulate that there are two dimensions, internal and external on which a person can be motivated to
11 respond without prejudice. This research re sulted in the development of two scales that measure these two sources of motiv ation to respond without prejudice: the internal motivation to respond without prejudice scale (IMS) and the external motivation to respond without prejudice scale (EMS). These two sources of motivation are separate from each other, distinguishing between Â“people who are primarily motivated by personal concerns to respond without prejudice (i.e., internally motivated) and those who are primarily motivated by concerns over how they might appear in the eyes of others (i .e., externally motivated)Â” (Plant & Devine, 1998, p. 812). Each source of motivation th eoretically falls on a distinctive continuum; a person could be high or low in both internal and ex ternal motivation. These motivational processes may strengthe n or weaken the effects of Aversive Racism and the present research atte mpted to clarify this relationship. Internal and external sources of motivation to respond without prejudice should have different eff ects on the expression and i nhibition of prejudice. Individuals high in internal motivation should likely be consistent in their attempts to control any expression of prej udice, whereas individuals hi gh in external motivation should be more influenced by the situational co ntext (Devine, Plant, and Blair, 2001). These two sources of motivation are re lated to implicit and explicit racial attitudes. Implicit racial pr ejudice is affected by the inte raction between internal and external motivation to respond without prejudice (Devine, Plant, Amodio, HarmonJones, & Vance, 2002). Devine and collea gues (2002) showed that individuals who scored high on the IMS reported lower explicit prejudice scores than individuals who
12 scored low on the IMS. More importantl y, individuals who scored high on the IMS scale and low on the EMS scale had lowe r scores on implicit racial prejudice measures. This effect was replicated with two different measures of implicit prejudice and also under conditions of c ognitive busyness, suggesting Â“that the low levels of race bias among high internal, low external individuals are not a result of effortful controlÂ” (Devine et al., 2002, p. 844). This relationship pattern ties into the Aversive Racism framework and demonstrat es that individuals high in internal motivation and low in external motivation ar e the most effective at inhibiting racial prejudice. Hypocrisy Induction as a M echanism to Reduce Prejudice A major goal of the present research wa s to explore strategies for reducing bias among Aversive Racists in the ratings of job applicants and s ubsequent selection. One strategy used to alter undesirable behavior is a hypocri sy manipulation. Hypocrisy manipulations have been used successfully to encourage condom use, recycling, and water conservation (Dic kerson, Thibodeau, Aronson, & Miller, 1992; Fried & Aronson, 1995; Stone, Wiegand, C ooper, & Aronson, 1997). In hypocrisyinduction experiments, individuals first make a public commitment requesting others not behave in a certain manner (e.g. non-prejud icial), and are then reminded of their own past behavior that cont radicts this request (e.g. re membering a time when they behaved in a prejudicial manner) (Dicke rson, Thibodeau, Aronson, & Miller, 1992). Hypocrisy procedures are similar to se lf-confrontation procedures (Rokeach, 1971) that are designed to create dissatisfaction in individuals by revealing inconsistent
13 information about the personÂ’s values and attitudes. Evidence shows that the hypocrisy effect is a form of cogn itive dissonance (Fried & Aronson, 1995), occurring when two contradictory thoughts held simultaneously create a motivated state to resolve this contradiction. Subse quently, this dissonance may result in a selfmodification of behavior. Research suppor ts this assertion. The induction of hypocrisy has been shown to change later re levant behaviors, whereas just reminding an individual of past failures to beha ve accordingly or making a public request advocating the behavior alone does not elic it change (Dickerson et al., 1992). A direct change in behavior as a result of the hypocrisy indu ction was preferred to selfaffirmative strategies designed to restore self-integrity (Stone, Wiegand, Cooper, & Aronson, 1997). Self-affirmative strategies al low individuals to avoid negative affect by focusing on positive self-attributes that are not related to the inconsistency make salient by the hypocrisy induction. In su mmary, the induction of hypocrisy involves reminding a person of past failures to a dhere to an ideal and making a public commitment to change, in an effort to elicit actual behavioral change in the individual. Hypocrisy induction can be very usef ul for studying racial attitudes and behavior. According to the Model of Dual Attitudes proposed by Wilson, Lindsey, and Schooler (2000), a person can have two se parate attitudes (one implicit and one explicit), that may not agree, toward an obj ect. Because implicit attitudes are outside of conscious control, a cont radiction between implicit and explicit attitudes does not create a state of cognitive dissonance. Therefore, a person who has implicit, but not
14 explicit prejudice, is not nece ssarily motivated to change, because he/she may not be aware of his/her implicit prejudice. A study by S on Hing, Li, and Zanna (2002) found that using hypocrisy procedures to reveal inconsistency between implicit and explicit attitudes reduced prejudicial behaviors in a sample of aversive racists, but it did not have an effect on i ndividuals whose implicit and explicit attitudes agreed. Hypocrisy makes salient this contradiction in attitudes. Using hypocrisy procedures has been shown to lower discriminatory behaviors. This has implications for selection decisions in the workplace. By using hypocrisy to expose internal contradictions, the present study will invest igate its effect on fairness in selection decisions. Present Study The present study compared a sample of Aversive Racists to a sample of truly non-prejudiced individuals on a job applicant-rating task. Both of these groups of individuals score low on explicit prejudice, but Aversive Racists are high on implicit prejudice, whereas truly nonprejudiced individuals are low on implicit prejudice. Participants were presente d with hypothetical resumes of four Â“applicantsÂ” for a management job. Selection ratings served as the dependent variable. Two of the applicants served as the critical comparis on. These two applicants had very similar qualifications, but systematically differed by race (except in the control group, where no race information will be provided). A pplicant qualifications were ambiguous for Black and White applicants. One applic ant had higher scholastic achievements (grade point average), while the other appl icant had greater work experience. Pre-
15 testing demonstrated that these two sets of qualifications are rated similarly when there were no race identifiers (Mean rating of the Â“superior scholasticÂ” applicant = 70.54, mean rating of Â“greater wo rk experienceÂ” applicant = 70.66, N = 92). In order to make sure that applicant qualifications was not a factor in the decision making process, race was applied to the two sets of applicant qualifications in a counterbalanced design. In this way, the ambiguity of the appli cantsÂ’ credentials was ensured. Hypocrisy was used as a means of altering selecti on decisions involving race. The main prediction was that the hypocrisy manipulation would have different effects on individuals who are characterized as Aversive Racists in comparison to individuals who are characterized as truly non-prejudiced individuals. Hypotheses The hypotheses focused on participant ch aracteristics (Aversive Racists versus truly non-prejudiced), situational ma nipulations (hypocrisy induction), and interactions between the two.
16 Hypothesis 1: Aversive R acists will rate Black app licants lower than White applicants, whereas truly non-prejudiced indivi duals will rate Black applicants similar to White applicants. Racism x RaceWhiteBlack Applicant RaceApplicant Ratings Aversive Racists Truly NonPrejudiced Figure 1 Hypothesis 1.
17 Hypothesis 2: Hypocrisy will alter applicant rati ngs so that participants in the hypocrisy condition will rate Black applican ts higher than White applicants, while individuals in the control condition will ra te Black and White applicants similarly. Hypocrisy x RaceWhiteBlack Applicant RaceApplicant Ratings Hypocrisy Control Figure 2. Hypothesis 2.
18 Hypothesis 3: Hypocrisy will only have an effect for Aversive Racists. Aversive Racists who are in the hypocrisy condition will rate Black applicants higher than White applicants, whereas Aversive Racist s in the control condition will rate White applicants higher than Black applicants. Aversive RacistsWhiteBlack RaceApplicant Ratings Hypocrisy Control Truly Non-PrejudicedWhiteBlack RaceApplicant Ratings Hypocrisy Control Figure 3 Hypothesis 3.
19 Hypothesis 4: External motivation to res pond without prejudice wi ll affect applicant ratings for Aversive Racists, but will not affect truly non-prejudiced individuals. Aversive Racists who are lower on external motivation will rate Black applicants lower than will Aversive Racists who are higher on external motivation to respond without prejudice. Aversive RacistsWhiteBlack Applicant RaceApplicant Ratings High External Motivation Low External Motivation Truly Non-PrejudicedWhiteBlack Applicant RaceApplicant Ratings High External Motivation Low External Motivation Figure 4. Hypothesis 4.
20 Hypothesis 5: External motivation to re spond without prejudice will moderate the effect of hypocrisy so that participants who score higher on external motivation will rate Black applicants higher than will pa rticipants who score lower on external motivation. HypocrisyWhiteBlack Applicant RaceApplicant Ratings High External Motivation Low External Motivation ControlWhiteBlack Applicant RaceApplicant Ratings High External Motivation Low External Motivation Figure 5 Hypothesis 5.
21 Method Participants Three-hundred twenty-four White undergra duates participat ed in the study, race being pre-selected through an online part icipant pool service. Seven participants identified themselves as non-Whites and were subsequently removed from the analyses. In addition, eight participants fa iled to fill out the R acial Climate Survey (the second part of the hypoc risy manipulation) and were removed from the analyses. A total of 309 participants were included in the final analyses. Participants were awarded extra credit in exchange for thei r contribution in the study. Typical of undergraduate samples, the mean age was 20.89 (SD = 4.49). In addition, the sample was 84.8% female (N = 262) and 15.2% male (N = 47). This study is concerned with comparing two groups of individuals, Aversive Racists and truly non-prejudiced. These two groups are bot h conceptualized as being low in explicit prejudice. Participants were given an explicit measure of racial prejudice (Modern Racism Scale), and those that scored above one standard deviation were not included in the anal yses that involved comparing Aversive Racists to truly non-prejudiced individuals.
22 Materials Explicit prejudice Explicit racial prejudice wa s measured using the Modern Racism Scale (MRS). This six-item s cale demonstrated acceptable internal reliability, alpha = .72. It al so correlated with the In ternal Motivation to Respond without Prejudice Scale (r = -.48), demonstrating some level of convergent validity. See Appendix A for the MRS. Implicit prejudice. The Implicit Association Te st (IAT) is a computerized test, which records reaction times that are used to measure implicit prejudice. The general logic of the test is that quicker reaction times indicate a stronger relationship between two concepts. For example, race can be divided up into two categories, White and Black. These two concepts are then paired with positive or negative words. A quicker reaction time for the pair ing (e.g. Black faces with negative words) demonstrates a stronger cogni tive association. Administra tion of the IAT was done in five stages: initial target-concept disc rimination, initial evaluative attribute discrimination, initial combined task, re versed target-concep t discrimination, and reversed combined task (Greenwald, McGhee, & Schwartz, 1998). Stage one (initial target-concept discrimination) involved being able to quickly categorize faces presented on a computer screen as Bl ack/African American or White/European American by pressing an appropriate key. Stage two (initial evaluative attribute discrimination) involved being able to categor ize words as positive or negative. This stage makes sure that individuals can differentiate between positive and negative words. Stage three (initial combined ta sk) involved combining the previous two
23 stages, for example White or positive versus Bl ack or negative. Half of the trials in this stage served as practice, and the other half were actually us ed in the analysis. Stage four (reversed target-concept discrimi nation) changed the si de (left or right key) that was assigned to Black or White, a nd the rest of stage f our replicated stage one. Stage five (reversed combined task) involved reversing the combination used in stage three, so continuing with the exampl e White would now be paired with negative while Black would be paired with positive. The rest of this stage replicated stage three, and the order of presentation was count erbalanced across multiple participants. Order of presentation does have a small eff ect on the expression of the IAT, so it is important to keep this in mind when interpreting the results. The conventional algorithm requires three steps in order to analyze the data. First, response latencies that are larger than 3,000 ms were recoded as 3,000 ms; response latencies that are smaller th an 300 ms were recoded as 300 ms as recommended by Greenwald et al. (1998). Second, the response latencies were log transformed before averaging them. Th ird, trials with error-latency (when a participant answers incorrectly to the stimulus) were included in the data (Greenwald, Nosek, & Banaji, in press). The IAT effect was measured by the difference between stage 3 and 5, excluding the practice portion of each stage. This final score was the difference between the initial combined task and the reversed combined task. In practical terms, higher scores on the IAT re flect a stronger associ ation of Black with negative and a stronger association of White with positive. The assumption is that
24 people who strongly associate Black with ne gative are demonstrating an implicit form of prejudice. The IAT has shown acceptable levels of reliability. Greenwald et al. (1998) also showed a significant m odified immediate test-retest reliability correlation of .46 for implicit racial attitudes. Test-retest re liability with a delay of 24 hours revealed a correlation of .65 (Dasgupta & Greenwald, 2001). Overall, the test-retest reliability performed by the authors has shown modest correlations between .27 and .85. Splithalf reliability correlatio n for the IAT were from r = .89 to r = .92 and the race IAT had an average internal consiste ncy of .57 (Greenwald & Nosek, 2001). In our study, the IAT data were proble matic and did not replicate previous findings of a pro-White bias in White part icipants (See Figure 6). Caution should be used when interpreting the results involving the IAT. The difference score computed for the IAT showed little bias or differen ce from zero (M = .06, SD = .46). The IAT is designed so that it is difficult for an i ndividual to control their responses, but it is possible that the partic ipants may have been able to al ter their responses in a socially desirable manner. Another explanation fo r the unexpected non-effect for the IAT could be the nature of the data collec tion. Four undergraduate experimenters collected the data. It is possible that th e experimenters did not run the IAT properly, but there is not eviden ce to indicate this. Internal and external motivat ion to respond without prejudice To measure motivation, two scales were used: the Internal Motivation to Respond Without Prejudice scale (IMS) and th e External Motivation to Respond Without Prejudice
25 scale (EMS) (Plant & Devine, 1998). The IMS and EMS are theoretically different constructs and should not necessarily relate with each other. In support of this, the IMS and EMS had a small, non-significant correlation (r = -.05) with each other. The IMS showed strong convergent validity, as it had high correlations in the expected direction, with the measure of explicit prejudice (e.g. r = -.48 for Modern Racism Scale). So, higher scores on the IMS correlate d with lower levels of racial prejudice. Because the EMS was designed to measure a personÂ’s desire to respond without prejudice because of apprehension abou t how others would view them if they responded in a prejudicial manner, the aut hors did not make any predictions about how the scale would relate to other measur es of racial prejudice. The EMS had a small, but significant co rrelation with MRS (r = .26). Both scales demonstrated good internal reliability; alpha = .79 for the EMS, and alpha = .87 for the IMS. See Appendix B for the IMS and EMS. See Ta ble 1 for a correlation matrix of the independent variables. Job applicant ratings In order to simulate a hiring selection scenario, participants were presented w ith four potential job applican ts. Only two of these job applicants are of interest in the analysis. The other two job applicants were used as anchors for the applicant ratings, and helped to mask the true intent of this study. For this purpose, one non-target applicant had extremely high credentials and the other non-target had extremely low credentials. The two job applicants of interest had more mixed qualifications. One target had a relatively high degree of work experience, and the other target had more education. Although th ese applicants had
26 distinct qualifications, the inte nt was to create resumes th at would be seen as roughly equal. However, the resume qualifications did have an effect on the no race control condition, such that participants favored the resume with higher experience. See Appendix C for job applicant rating task materials, including the resumes. Participants rated the app licants using a global 1-100 sc ale. In addition to the global ratings, participan ts also rated applicants on three questions: This is a suitable candidate for the position of Intermediary Supervisor; I w ould offer this candidate the position of Intermediary Supervisor ; and I would offer this candidate a position somewhere in the organization These ratings used Likert-type responses on a sevenpoint scale and were combined do form a composite rating. The composite rating demonstrated acceptable internal consistency, alpha > .85. See Table 2 for a correlation matrix of th e dependent variables. Design Individuals who scored hi gh in explicit prejudice (two standard deviations above the mean) were eliminated from the an alyses that involved the Aversive Racist and truly non-prejudiced dis tinction, while the remaining individuals who scored low in explicit prejudice were split into two groups (Aversive Racists and truly nonprejudiced). Eliminating those high in explicit prejudice left 256 participants. The participants were randomly assigned to ei ther the hypocrisy condi tion or the control condition. Next, one of three different versions of job applicant qualifications was given to each participant: two of the sets of resumes included applicant photos (one Black and one White) to make race a sa lient characteristic of each applicant
27 (counterbalancing across applicant resumes) while the third condition did not have any pictures in order to serve as a race ne utral control. In the two race conditions, the two non-target applicants were White. Ge nder was held constant (only male job applicants) in order to simplify the design, but should be looked at in future research. Participants were asked to examine and rate four job applicants. The applicant ratings served as the dependent variable. Only the two mid-qualification applicants were used in the analysis, while the other two distracter applicants were intended to be used as comparison anchors for the applicant ratings such that one applicant had superior qualifications and the other had extremely low credentials. Applicant materials were systematically a ltered so that there were three between group conditions. These three initial variables constitute d a 2 (Aversive Racists versus truly non-prejudiced) x 2 (hypocrisy induction versus control) x 3 (applicant race) factorial design. The fourth independent variable external motivation to respond without prejudice, was hypothesized to be a moderato r in this process a nd was analyzed using ANCOVA. Procedure Participants were presen ted with three stages in the study. Each stage was presented to the participants as being inde pendent experiments in order to mask the true intent of the research due to the sensitiv e nature of prejudice. In the first stage, participants were given the explicit prej udice scale (MRS), the motivation scales (IMS and EMS), and the Modern Sexism Scale (to serve as a distracter). This stage
28 was titled Â“Social AttitudesÂ” and was desc ribed as an experime nt being conducted by a social psychologist in the department. The second stage involved the computeri zed administration of the IAT, and was described as a test of peopleÂ’s automatic cognitive associations of various categories being done by a Cognitive psychologist in the psychology department. The description was intended to prevent par ticipants from realizing later tasks were related to the IAT. The third section was described as a miscellaneous questionnaire packet, which consisted of the hypocrisy manipula tion in the guise of a public service study being conducted by the Diversity Committ ee of Graduate Researchers. The hypocrisy procedure is very similar to the one used by Son Hing, Li, and Zanna (2002). In the hypocrisy condition, participan ts were asked to write a persuasive essay discussing the message that racial prejudice and disc rimination are still problems that exist in our society. Particip ants were told that excerpts from these essays would be eligible fo r inclusion in pamphlets aime d at high school students for a Â“Racial Equity Forum.Â” See Appendix D for specific instructions. Next, the participants completed a quick cognitive filler task (e.g. a word descramble, see Appendix D) and then filled out a racial climate survey that asks the following: The psychology department is interest ed in understanding issues of race relations in our culture. Specifically, we are interested in developing some scenarios based on the actual experiences of people like yourself. We would like you to take a few moments to writ e about examples from your past.
29 Briefly write about 2 instances in which you reacted negatively in some way toward an African American Â– for instance, treating someone in a prejudiced manner, having a negative thought or attitu de, having a negative job-related or school-related experience, and so on. K eep in mind that this information will be kept confidential and anonymous. Participants then completed a revised version of the Positive and Negative Affect Scale (PANAS; Watson, Clark, & Telleg en, 1988) aimed at measuring feelings of discomfort and guilt. See Appendix E. Th is served as a check to make sure that the hypocrisy procedure worked. Previous research has shown that hypocrisy is a form of dissonance, which leads to negativ e feelings such as guilt (Fried & Aronson, 1995). Hypocrisy only results when an indi vidual first publicly a dvocates a particular opinion, and then remembers a time in whic h they behaved contrary to that opinion (Stone, Aronson, Crain, Winslow, & Frie d, 1994). The control condition followed these exact proceedings except the persuasive message advocated against smoking in order to ensure that just remembering a time in which a person behaved negatively toward an African American does not alone elicit guilty feelings. The Â“Business Hiring DecisionÂ” task was also included in stage three. This simulated job selection scen ario was described as an experiment being conducted by an industrial/organizational ps ychologist in the departme nt who is interested in selection decisions. Again, par ticipants were asked to rate four job applicants. Only two of these applicants were used in the analysis (Candidates B and C in Appendix C materials). Once a rating was obtained for each of the job applicants, the participants
30 were asked to give explanati ons for their ratings. This served as a check to ensure that the two ambiguous applicants did not differ on some important qualification other than race. Analyses To test the first three hypotheses, a 2 (aversive racists vs. truly nonprejudiced) x 2 (hypocrisy versus contro l) x 3 (applicant race) factorial ANOVA design was used. Although implicit and ex plicit prejudice scores are c ontinuous, they needed to be treated as categorical in order to create the Aver sive Racist and truly nonprejudiced categories, which are interactions between the implicit and explicit measures. If left as continuous variables, in terpretation of interaction scores would be problematic (that is, a high score on implicit and a low score on explicit, or vice versa, would result in th e same interaction score). For hypotheses four and five, ANCOVAs were performed to test for the moderating effects of external motivati on to respond without prejudice, because this variable is being treated as continuous. A three-way interaction determines whether external motivation moderates the interactions be tween Aversive Racism and applicant ratings, and between hypocrisy induc tion and applicant ratings.
31 Results Hypotheses Overall, the Black applicant (M = 81.03, SD = 12.29) was rated higher than the White applicant (M = 76.83, SD = 14.20), F (1, 205) = 17.89, p <.001 on a global 1 to 100 rating. The two candidates were ge nerally rated quite highly, but there were a few extreme outliers who gave unusually lo w ratings to the applicants. Twelve participants rated applicants below 50, a score roughly two standard deviations below the mean. Scores below 50 were rescored to 50 in order to lessen the effect of these outliers while keeping these individuals in the sample. When reanalyzed using the corrected scores, race still had a main effect on ratings such that the Black applicant (M = 81.44, SD = 10.62) was rated higher th an the White applicant (M = 77.60, SD = 11.65), F (1, 205) = 19.37, p <.001. In addition, using th e mean Likert composite revealed a similar main effect for race: the Black applicant (M = 5.14, SD = 1.03) was rated higher than the White applicant (M = 4.49, SD = 1.15), F (1, 205) = 39.85, p <.001. All analyses reported below use the corrected global ratings. Hypothesis 1 Aversive Racists will rate Bl ack applicants lower than White applicants, whereas truly non-prejudiced indivi duals will rate Black applicants similar to White applicants.
32 There was a significant interaction betw een applicant race and level of racism (Aversive Racists versus truly non-prejudiced individuals), F (1, 163) = 4.16, p = .04. Contrary to Hypothesis 1, the pattern of this interaction revealed that both Aversive Racists and truly non-prejudiced individuals rated the Black applicant (for Aversive Racists, M = 84.05, SD = 8.56; for truly non-prejudiced, M = 80.64, SD = 11.80) higher than the White applican t (for Aversive Racists, M = 76.79, SD = 11.65; for truly non-prejudiced, M = 77.15, SD = 12.40), but this effect was greater for the Aversive Racists (F (1, 83) = 29.49, p < .001) than for tr uly non-prejudiced (F (1, 80) = 7.53, p = .007). However, this interaction was not significant when examining the Likert composite average, (F (1, 163) = 1.52, p = .22). In addition, the data from the Implicit Association Test (IAT) did not rep licate previous findings that found a proWhite bias (see Figure 6). Interpretation of the results involving the IAT should be made with caution. Other individual differe nce measures, such as explicit prejudice (the Modern Racism Scale), are addressed in the additional analyses section. These measures may be more appropriate for interpretation. Hypothesis 2 Hypocrisy will alter a pplicant ratings so that participants in the hypocrisy condition will rate Black applican ts higher than White applicants, while individuals in the control condition will ra te Black and White applicants similarly. In support of Hypothesis 2, the inducti on of hypocrisy produced a significant interaction with applicant race, using both the corrected global (50-100) ratings, F (1, 204) = 4.04, p =.046, and the Likert composite ratings, F (1, 204) = 4.36, p =.038. The pattern of this inte raction was such that inducing hypocrisy increased the difference
33 between participantsÂ’ ratings of the Black and White appl icant (see Table 1). When participants were not induced with hypoc risy, they rated the Black candidate 2.11 points higher than the White candidate (F (1, 103) = 3.34, p = .07) on the global ratings. However, this difference betw een the candidates increased to 5.60 (F (1, 101) = 18.78, p < .001) when they were made to feel hypocritical. This pattern was also found when analyzing the Like rt composite. When participants were not induced with hypocrisy, they rated the Black candidate .44 points hi gher (on a 7 point scale) than the White candidate (F (1, 103) = 10.38, p = .002) on the Likert composite. This difference increased to .87 when the partic ipants were made to feel hypocritical (F (1, 101) = 32.19, p < .001). Hypothesis 3. Hypocrisy will only have an effect for Aversive Racists. Aversive Racists who are in the hypocrisy c ondition will rate Black applicants higher than White applicants, whereas Aversive Racists in the contro l condition will rate White applicants higher than Black applicants. Contrary to Hypothesis 3, the 3-way in teraction between racism, applicant race, and hypocrisy was not significant using the corrected global ratings (F (1, 161) = .089, p = .77) or the Likert composite (F (1, 161) = .755, p = .39). The effect of the hypocrisy manipulation did not differ for Av ersive Racists in comparison to truly non-prejudiced individuals. Hypothesis 4. External motivation to respond without prejudice will affect applicant ratings for Aversive Racists, but will not affect truly non-prejudiced individuals. Aversive Racists who are lo wer on external motivation will rate Black
34 applicants lower than will Aversive Raci sts who are higher on ex ternal motivation to respond without prejudice. Contrary to Hypothesis 4, the 3-way in teraction between racism, applicant race, and external motivation to respond without prejudice was non-significant using the corrected global ratings (F (1, 161) = .051, p = .82) and the Likert composite (F (1, 161) = .100, p = .75). External motivation to respond without prejudice did not moderate the effect of raci sm on the applicant ratings. Hypothesis 5. External motivation to respond without prejudice will moderate the effect of hypocrisy so th at participants who score hi gher on external motivation will rate Black applicants higher than will participants who score lower on external motivation. Contrary to Hypothesis 5, the 3-way in teraction between hyp ocrisy, applicant race, and external motivation to respond without prejudice was non-significant using the corrected global ratings (F (1, 202) = .048, p = .83) or Likert composite (F (1, 202) = .247, p = .62). External motivation to res pond without prejudice did not moderate the effect of hypocrisy on the applicant ratings. Additional Analyses Hypocrisy Effect: Mediation by Emotions. Hypocrisy also had an effect on the emotional state of the participants. After being introduced to the hypocrisy manipulation, participants re ported their current mood by responding to twenty mood items, two of which were included to measur e feelings of guilt specifically: guilty and ashamed. Participants in the hypocrisy condition reported feeling more guilty (M =
35 2.04, SD = 1.51) than the participants in the control condition (M = 1.69, SD = 1.33), F(1, 308) = 4.72, p=.03, and more ashamed (M = 1.85, SD = 1.32) than the participants in the cont rol condition (M = 1.59, SD = 1.10), F(1, 307) = 3.53, p=.061. Of the eighteen remaining mood rati ngs, no other significant differences emerged (all p s > .05), with two exceptions. Partic ipants in the hypocrisy condition reported feeling less excited (M = 2.90, SD = 1.38) than particip ants in the control condition (M = 3.23, SD = 1.53), F (1, 308) = 4.00, p =.046, and more upset (M = 2.46, SD = 1.67) than participants in the control condition (M = 2.05, SD = 1.34), F (1, 308) = 5.65, p =.018. Because only feelings of guilt and shame are of theoretical interest, guilt and shame ratings were combined ( = .84). A four-step mediat ion analysis was run with this composite variable following the pr ocedure outlined by Baron and Kenny (1986). Please see Figure 7 for a graphical depict ion. The dependent variable was the difference between the corrected Black a nd White applicant ratings, with higher scores indicating a preference for the Black candidate. In step one, regressing the Black-White difference score on hypocri sy was statistically significant, = .139, t (205) = 1.74, p = .046. In step two, regressing the guilt composite on hypocrisy was statistically significant, = .123, t (205) = 2.17, p = .031. In step three, regressing the Black-White difference score on both hypocrisy and the guilt composite rendered guilt marginally significant, =.131, t (205) = 1.87, p = .063, but not hypocrisy, = .116, t(205) = 1.66, p = .10. Therefore, the effect of hypocrisy on the difference
36 between the corrected ratings of Black and White applicants is mediated by feelings of guilt. Individual Differences. There was a significant in teraction between applicant race and scores on the Modern Racism Scal e (MRS) on the corrected ratings of the job applicants, F(1, 204) = 16.26, p < .001, such that those higher in racism rated the Black applicant lower than those lower in racism (r = -.27, p < .001) while the MRS was not related to the White applicant rati ng (r = .05, p = .52). A similar interaction resulted when the composite Likert ratings of the applicants were used, F(1, 204) = 13.03, p < .001. There was also a significant interacti on between applicant race and scores on the Internal Motivation to Respond without Prejudice Scale (IMS) on the corrected ratings of the job applicants, F (1, 204) = 8.97, p = .003, such that those higher on the IMS scored the Black applicant higher th an those lower in internal motivation (r = .22, p = .001) while the IMS was not relate d to the White applicant rating (r = -.02, p = .80). A similar interaction resulted wh en the composite Likert ratings of the applicants was used, F (1, 204) = 12.20, p = .001. A significant interaction was found betw een applicant race and scores on the External Motivation to Respond without Prej udice Scale (EMS) on the Likert ratings of the job applicants, F (1, 204) = 3.90, p < .05, such that those higher on the EMS scored White applicants higher than those lower on external motivation (r = .16, p = .03) while the EMS was not related to the Black applicant rating (r = -.03, p = .73).
37 But this interaction was not found when exam ining the corrected job applicant scores, F (1, 204) = 1.32, p = .25. There were no significant three-way interactions between applicant race, hypocrisy, and any of the indivi dual difference measures (all ps > .05). This refutes the idea that individual differences mode rate the interaction between hypocrisy and applicant race on the ratings of the job applicants.
38 Discussion This study examined how Whites' subtle prejudices could impact selection decisions in a hiring context. Hypocrisy wa s examined as a means to alter selection decisions, so that the ratings of a Black applicant would be increased. It was predicted that Aversive Racists, those indi viduals who show little overt prejudice but who may harbor automatic and less controllabl e bias, would be espe cially affected by a hypocrisy manipulation. The analyses that involved compar ing Aversive Raci sts and truly nonprejudiced individuals shoul d be interpreted with caution. The IAT data were somewhat problematic in that they did not show the anticipated pro-White bias. When looking at the results for hypothesis one (Aversive Racists will rate Black applicants lower than White applicants, wh ereas truly non-prejudi ced individuals will rate Black applicants similar to White appl icants), it appears that applicant race was more of an important factor for individua ls categorized as Av ersive Racists. Although the results reveal a bias in favor of the Black candidate, this was only significant for the Aversive Racists. Aver sive Racists are assu med to harbor some level of subtle prejudice that is automatic and implicit; a different interpretation could be that these individuals ar e the most uncomfortable in regards to race, or wrestle with their own ambivalence with race. When race is salient, they may try to
39 overcompensate for these negative feelings and actually demonstrate an out-group bias. In this study, Aversive Racists could have been prim ed by the emphasis of race in the social attitudes qu estionnaire packet (MRS, IMS, and EMS). Ideally, these prejudice measures could have been collect ed after the dependent variable (hiring decisions) and the IAT. Because of th e hypocrisy manipulation, this was not an option. The manipulation would have probably influenced those scores; consequently, the individual difference meas ures needed to precede the dependent variable. Still, the IAT us ed to create these two groups (Aversive Racists and truly non-prejudiced) does not lead to a clear in terpretation of this interaction. Other analyses that did not utilize the IAT appear to be more interpretable. When group categorization (Aversive Raci sts or truly non-prejudiced) was ignored, the hypocrisy effect was still found. Regardless of indi vidual differences in prejudice, the hypocrisy manipulation caused people to rate the Black applicant higher than the White applicant. The results support the idea that White people can be made to feel racial anxiety, which is made salient through th e hypocrisy condition. This study provides evidence in support of hypocrisy as an e ffective strategy in altering selection decisions involving race. Unlike previous research, these findings were not dependent upon individual differences (S ong Hing, Li, & Zanna, 2002). Instead this depiction described the entire sample of co llege students. The induction of hypocrisy increased the advantage given to the Black applicant relative to the White applicant compared to participants not made to feel hypocritical.
40 The hypocrisy effect appears to be driv en by emotions, specifically guilt and shame. The hypocrisy manipulation created hi gher levels of guilt and shame than did the control condition. A mediation analysis showed that the hypocrisy effect was mediated (at least partially) by these two emotions. This result replicates previous research that has shown hypocrisy to be e ffective in changing behavior (Dickerson, Thibodeau, Aronson, & Miller, 1992; Fr ied & Aronson, 1995; Stone, Wiegand, Cooper, & Aronson, 1997), and improves upon them by showing that guilt and shame are the specific emotions that mediate the hypocrisy effect in th e context of a race based decision. Further research is needed in order to examine whether or not guilt and shame, with regard to race, are distinct emotions. The hypocrisy manipulation alters se lection decisions, but this does not necessarily mean that these decisions are more objective. The goal of the hypocrisy manipulation was to change White particip antsÂ’ ratings of Black job applicants. Whether this type of a procedure would be useful outside of the lab setting is debatable. Further research is needed to determine whether or not these findings can be replicated in an applied setting with an older population. More importantly, research needs to determine what other consequences hypocrisy may have on individuals. It appears that guilt and shame are the emotions that are involved in this process. Using these emoti ons as a strategy to change behavior may or may not be appropriate in the corporate environment and could potentially have negative side effects for long-term behavior. Further research is necessary to resolve these concerns about the use of hypocrisy technique outside of the lab.
41 Individual differences, such as the m odern racism and internal motivation to respond without prejudice, were found to predic t a significant difference in the ratings of the White and Black applican ts regardless of whether the participants were in the hypocrisy or control condition. The modern racism scale and internal motivation to respond without prejudice scale are highl y (negatively) co rrelated, but are theoretically distinct constructs. Individua ls who were high in modern racism rated the Black applicant lower than those who we re low in modern racism. Individuals who were high in internal motivation to respond without prejudice rated the Black applicant higher than those w ho were lower in internal motivation. Neither of these constructs was related to the ratings of th e White applicant. It appears that these individual differences were re lated to out-group inflation a nd did not affect the ratings of the in-group member. Add itional research should addres s how this process works. Are people fairly set in their assessment of in-group versus out-group members? This study gives us a preliminary guess that ther e is a difference between in-group and outgroup evaluations. The results demonstrating an overall ma in effect for applicant race should be interpreted with caution. Due to the se nsitive nature of the study (or to demand characteristics of the study), it seems prem ature to conclude that Whites in general inflate out-group ratings. Some scholars have used the term White Guilt to describe such a phenomenon in which a White indivi dual actually favors an out-group member (usually in terms of racial categorization), which leads to preferential treatment for the out-group member (Steele, 2002; Sw im, & Miller, 1999; Iyer, Leach, & Crosby,
42 2003). Limitations in the study design prev ent this conclusion. The participants answered sensitive questions pertaining to race for about ten minutes before completing the IAT task. Even though these two parts of the study were said to be unrelated to the final section in which the selection task was performed, it is conceivable that the participan ts did not believe the experime nter or that this material primed the participants resulting in the in flation of the Black a pplicant ratings. In fact, follow-up data show a reverse eff ect. When only the selection task was performed (no prejudice measures or hypocri sy manipulation), part icipants rated the White applicant higher than the Black applican t. It seems that Whites can be made to feel guilt and shame due to their race, but this is not necessarily the social norm for Whites in absence of these particular cues. Further research is needed in order to resolve this issue and determine what types of cues elicit such be havior and in what context. Improvements in the study design could be made in order to conceal the true nature of the research by posting individua l difference (MRS, IMS, EMS, etc.) online as a prerequisite for participation. Th e disadvantage is that the IAT would be extremely difficult to administer given the da ta collection procedures at USF. Future research should tease apart the independe nt effects of hypocrisy on shame versus guilt. It seems possible that guilt (an inward-focused emotion) and shame (an outward-focused emotion) may have di ffering effects on whether the individual deflates the ratings of their in-group or inflates their ra tings of an out-group. In addition, the effects of hypocrisy could be influenced by racial identification. For
43 instance, Whites who highly identify with thei r race may be less likely to be affected by hypocrisy, while Whites who do not strongly identify with their race may be more likely to be affected by hypocrisy. Overa ll, this study has replicated and extended previous research in that hypocrisy inducti on was shown to have an effect on Whites' ratings of Blacks, and this induction was me diated by feelings of guilt and shame.
44 References Adorno, T., Frenkel-Brunswik, E., Levins on, D., & Sanford, R. N. (1950). The authoritarian personality New York: Harper. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Co nceptual, strategic and statistical considerations. Journal of Pers onality and Social Psychology, 51, 1173-1182. Bosson, J. K., Swann, W. B., & Pennebaker, J. W. (2000). Stalking the perfect measure of implicit self-esteem: The bli nd men and the elephant revisited? Journal of Personality and Social Psychology, 79 631-643. Brigham, J. C. (1993). College students' raci al attitudes. Journal of Applied Social Psychology, 23 1933-1967. Cohen, J. (1988). Statistical power an alysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Crandall, C. S., Eshleman, A. & OÂ’Brien, L. (2002). Social norms and the expression and suppression of prejudice: The struggl e for internalization. Journal of Personality and Social Psychology, 82 359-378. Dasgupta, N., & Greenwald, A. G. (2001). On the malleability of automatic attitudes: Combating automatic with images of admired and disliked individuals. Journal of Personality and Social Psychology, 81 800-814.
45 Devine, P. G., Plant, E. A., Amodio, D. M., Harmon-Jones, E., & Vance, S. L. (2002). The regulation of explicit and implicit race bi as: The role of motivations to respond without prejudice. Journal of Personality and Social Psychology, 82 835-848. Devine, P. G., Plant, E. A., Blair, I. V. (2001). Classic and cont emporary analyses of racial prejudice. In R. Brown & S. Gaertner (Eds.), Blackwell Handbook of Social Psychology: Intergroup Processes. (pp. 198-217). Blackwell Publishing, Malden: MA. Dickerson, C. A., Thibodeau, R., Aronson, E ., & Miller, D. (1992). Using cognitive dissonance to encourage water conser vation. Journal of Applied Social Psychology, 22 841-854. Dovidio, J. F. (2001). On the nature of contemporary prejudice: The third wave. Journal of Social Issues, 57 829-849. Dovidio, J. F., & Gaernter, S. L. (1986). The aversive form of racism. In J. F. Dovidio & S.L. Gaertner (Eds.), Pr ejudice, discrimination, and racism. (pp. 61-89). San Diego, CA: Academic Press. Dovidio, J. F., & Gaertner, S. L. (1998). On the nature of contemporary prejudice: The causes, consequences, and challenges of aversive racism. In J. Eberhardt & S. T. Fiske (Eds.), Confronting racism: The problem and the response. (pp. 3-32). Newbury Park, CA: Sage. Dovidio, J. F., & Gaertner, S. L. (2000). Aversive racism and selection decisions: 1989 and 1999 Psychological Science, 11 315-319.
46 Dovidio, J. F., Gaertner, S. L., Anastasio, P. A, & Sanitioso, R. (1992). Cognitive and motivational bases of bias: Implications of aversive racism for attitudes toward Hispanics. In S. B. Knouse, P. Rosenfeld, & A. L. Culbertson (Eds.), Hispanics in the workplace. (pp. 75-106). Newbury Park, CA: Sage Publications. Dovidio, J. F., Kawakami, K., & Beach, K. R. (2001). Implicit and explicit attitudes: Examination of the relationship between measures of intergroup bias. In R. Brown & S. Gaertner (Eds.), Black well Handbook of Social Psychology: Intergroup Processes (pp. 198-217). Blackwell Publishing, Malden: MA. Dovidio, J. F., Kawakami, K., & Gaertner S. L. (2002). Implicit and explicit prejudice and interracial interaction. Journal of Personality and Social Psychology, 82 62-68. Dovidio, J. F., Kawakami, K., Johnson, C ., Johnson, B., & Howard, A. (1997). On the nature of prejudice: Automatic a nd controlled processe s. Journal of Experimental Social Psychology, 33 510-540. Fazio, R. H., Jackson, J. R., Dunton, B. C., & Williams, C. J. (1995). Variability in automatic activation as an unobtrusive meas ure of racial att itudes: A bona fide pipeline? Journal of Persona lity and Social Psychology, 69 1013-1027. Fried, C. B., & Aronson, E. (1995). H ypocrisy, misattribution, and dissonance reduction. Personality and Social Psychology Bulletin, 21 925-933.
47 Gaertner, S. L., & Dovidio, J. F. (2000). Aversive racism and intergroup biases. In S. L. Gaertner & J. F. Dovidio (Eds .), Reducing intergroup bias: The common ingroup identity model. (pp. 13-31). Ann Arbor, MI: Sheridan Books. Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: Attitudes, selfesteem, and stereotypes. Psychological Review, 102 4-27. Greenwald, A. G., & Nosek, B. A. (2001). H ealth of the Implicit Association Test at age 3. Zeitschrift fur Experimentelle Psychologie, 48 85-93. Greenwald, A. G., McGhee, D. E., & Sc hwartz, J. L. K. (1998). Measuring individual differences in implicit cogn ition: The Implicit Association Test. Journal of Personality and Social Psychology, 74 1464-1480. Greenwald, A. G., Nosek, B. A., & Banaji, M. R. (in press). Understanding and Using the Implicit Association Test: 1. An Improved Scoring Algorithm. Journal of Personality and Social Psychology Iyer, A., Leach, C. W., & Crosby, F. J. ( 2003). White guilt and racial compensation: The benefits and limits of self-focus Personality and Social Psychology Bulletin, 29 117-129. 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. McConahay, J. B. (1986). Modern racism ambivalence, and the modern racism scale. In J. F. Dovidio & S. L. Gaer tner (Eds.), Prejudice, discrimination, and racism. (pp. 91-125). Orlando, FL: Academic.
48 Nosek, B. A., Banaji, M. R., & Greenwald, A. G. (2002). Harvesting implicit group attitudes and beliefs from a demons tration website. Group Dynamics, 6 101115. Pettigrew, T. F. (1959). Regional differences in anti-negro prejudice Journal of Abnormal and Social Psychology, 59 28-36. Plant, E. A., & Devine, P. G. (1998). In ternal and external motivation to respond without prejudice. Journal of Pe rsonality and Social Psychology, 75 811-832. Rokeach, M. (1971). Long-range experimental modification of values, attitudes, and behavior. American Psychologist, 26 453-459. Schuman, H. & Krysan, M. ( 1999). A historical note on Wh ites' beliefs about racial inequality. American Sociological Review, 64 847-855. Sears, O. (1988). Symbolic racism. In P. A. Katz & D. A. Taylor (Eds.), Eliminating racism: Profiles in controversy (pp. 53-84). New York: Plenum Press. Son Hing, L. S., Li, W., & Zanna, M. P. (2002). Inducing hypocrisy to reduce prejudicial responses among aversive racists. Journal of Experimental Social Psychology, 38 71-78. Steele, S. (2002). The age of white guilt and the disappearance of the black individual. New York: HarperPerennial. Stone, J., Aronson, E., Crain, A. L., Winslow, M. P., & Fried, C. B. (1994). Inducing hypocrisy as a means of encouraging young adults to use condoms. Personality and Social Psychology Bulletin, 20 116-128.
49 Stone, J., Wiegand, A. W., Cooper, J., & Aronson, E. (1997). When exemplification fails: Hypocrisy and the motive for self-i ntegrity. Journal of Personality and Social Psychology, 72 54-65. Swim, J. K., & Miller, D. L. (1999). White guilt: Its antecedents and consequences for attitudes toward affirmative actio n. Personality and Social Psychology Bulletin, 25, 500-514. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54 1063-1070. Wilson, T. D., Lindsey, S., & Schooler, T. Y. (2000). A model of dual attitudes. Psychological Review, 107 101-126. 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, 72 262-274. Wittenbrink, B., Judd, C. M., & Park, B. (2001) Spontaneous prejudice in context: Variability in automatically activated at titudes. Journal of Personality and Social Psychology, 81 815-827.
51 Appendix A: Modern Racism Scale PLEASE USE THE SCALE BELOW TO RESPOND TO THE ITEMS THAT FOLLOW BY WRITING A NUMBER BETW EEN 1-9 IN THE BLANK BEFORE EACH STATEMENT. 1 2 3 4 5 6 7 8 9 Strongly Disagree Neutral Strongly Agree ____ 1. It is easy to understand the a nger of Black people in America. ____ 2. Blacks are getting too demanding in their push for equal rights. ____ 3. Over the past few years, Blacks have gotten more economically than they deserve. ____ 4. Over the past few years, the govern ment and news media have shown more respect to Blacks than they deserve. ____ 5. Blacks should not push themselv es where they are not wanted. ____ 6. Discrimination against Blacks is no longer a problem in the United States.
52 Appendix B: Internal/External Moti vation to Respond without Prejudice Instructions : The following questions concern various reasons or motivations people might have for trying to respond in nonprejud iced ways towards Black people. Some of the reasons reflect internal-personal motivatio ns whereas others reflect more externalsocial motivations. Of course, people may be motivated for both internal and external reasons; we want to emphasize that neither t ype of motivation is by definition better than the other. In addition, we want to be clear that we are not evaluating you or your individual responses. All your responses will be completely confidential. We are simply trying to get an idea of the types of motiv ations that students in general have for responding in nonprejudiced ways. If we are to learn anything useful, it is important that you respond to each of the questions openly and honestly. Please give your response according to the scale below by writing a number fr om 1-9 in the space to the left of each statement: 1 2 3 45 6789 Strongly Disagre e Neutral Strongly Agree ____1. Because of todayÂ’s PC (politically correct) stan dards I try to appear nonprejudiced toward Black people. ____2. I attempt to act in nonprejudiced ways toward Black people because it is personally important to me. ____3. I attempt to appear nonprejudiced towa rd Black people in order to avoid disapproval from others. ____4. I am personally motivated by my beliefs to be nonprejudiced toward Black people. ____5. I try to act nonprejudiced toward Bl ack people because of pressure from others. ____6. I try to hide any negative thoughts ab out Black people in order to avoid negative reactions from others. ____7. According to my personal values, using stereotypes about Black people is OK. ____8. If I acted prejudiced toward Black peopl e, I would be concerned that others would be angry with me. ____9. Because of my personal values, I believe that using stereotypes about Black people is wrong. ____10. Being nonprejudiced toward Black peop le is important to my self-concept.
53 Appendix C Business Hiring Decision When businesses hire new employees, many f actors go into their decisions. They must consider multiple pieces of information, and decide how much weight to give to each. Often, they have little more to base their decision on than a stack of resumes or some brief interviews. Learning about the types of information that employers find important in the decision-making process is an important goal. Industrial-Organizational psychologists at USF are interested in le arning what criteria college students think are important when making hiring decisions and for predicting later success. For this task, we would like you to imagine that you are on the Board of Executives at a large corporation. Your job is to select the applicant who will be hired for the job of Senior Supervisor This is a mid-level job in your corporation that requires a highly competent person. The only formal job requ irement is that the job applicant has a Bachelors degree in Business Administration. On the next page, you will see summaries of 4 applicants resumes. Then on the following 4 pages, you will find ratings (1 page for each candidate). Your task is to review the qualifications of each applicant and make some ratings (assume this is the only information available to you).
54 Appendix C (Continued) CANDIDATE A Stephen R. Williamson Wallace University Education: Bachelor of Business Administration Masters of Business Administration GPA: 3.93/ 4.00 Experience: Program Assistant (1992-1994) Event Specialist (1994-1998) Design Director (1998-2000) Administration Manager (2000-2003) Skills: Microsoft Office Systems Development Visual Basic Macromedia Manuscript Production Performance Appraisal CANDIDATE B Jason L. Atkinson Spencer University Education: Bachelor of Business Administration GPA: 3.78/ 4.00 Experience: Director Assistant (2002-2003) Skills: Marketing Development Macromedia Microsoft Office CANDIDATE C Raymond D. Stevens Bellevue University Education: Bachelor of Business Administration GPA: 3.08/ 4.00 Experience: Design Specialist (1995-1997) Project Director (1997-1999) Assistant Manager (1999-2003) Skills: Marketing Development Visual Basic Design Microsoft Office CANDIDATE D Robert T. Jackson Carnot College Education: Bachelor of Business Administration GPA: 2.70/ 4.00 Experience: Graphic Engineer (1999-2003) Skills: Microsoft Office Internet Design
55 Appendix C (Continued) CANDIDATE A Stephen R. Williamson Wallace University Education: Bachelor of Business Administration Masters of Business Administration GPA: 3.93/ 4.00 Experience: Program Assistant (1992-1994) Event Specialist (1994-1998) Design Director (1998-2000) Administration Manager (2000-2003) Skills: Microsoft Office Systems Development Visual Basic Macromedia Manuscript Production Performance Appraisal CANDIDATE B Jason L. Atkinson Spencer University Education: Bachelor of Business Administration GPA: 3.78/ 4.00 Experience: Director Assistant (2002-2003) Skills: Marketing Development Macromedia Microsoft Office CANDIDATE C Raymond D. Stevens Bellevue University Education: Bachelor of Business Administration GPA: 3.08/ 4.00 Experience: Design Specialist (1995-1997) Project Director (1997-1999) Assistant Manager (1999-2003) Skills: Marketing Development Visual Basic Design Microsoft Office CANDIDATE D Robert T. Jackson Carnot College Education: Bachelor of Business Administration GPA: 2.70/ 4.00 Experience: Graphic Engineer (1999-2003) Skills: Microsoft Office Internet Design
56 Appendix C (Continued) CANDIDATE A Stephen R. Williamson Wallace University Education: Bachelor of Business Administration Masters of Business Administration GPA: 3.93/ 4.00 Experience: Program Assistant (1992-1994) Event Specialist (1994-1998) Design Director (1998-2000) Administration Manager (2000-2003) Skills: Microsoft Office Systems Development Visual Basic Macromedia Manuscript Production Performance Appraisal CANDIDATE B Jason L. Atkinson Spencer University Education: Bachelor of Business Administration GPA: 3.78/ 4.00 Experience: Director Assistant (2002-2003) Skills: Marketing Development Macromedia Microsoft Office CANDIDATE C Raymond D. Stevens Bellevue University Education: Bachelor of Business Administration GPA: 3.08/ 4.00 Experience: Design Specialist (1995-1997) Project Director (1997-1999) Assistant Manager (1999-2003) Skills: Marketing Development Visual Basic Design Microsoft Office CANDIDATE D Robert T. Jackson Carnot College Education: Bachelor of Business Administration GPA: 2.70/ 4.00 Experience: Graphic Engineer (1999-2003) Skills: Microsoft Office Internet Design
57 Appendix C (continued) Selection Page Based on the summaries of the 4 applicants Â’ resumes on the previous page, answer the following questions: For Candidate A: 1) Please rate the applicantÂ’s overall cred entials based on the information given (1-100, 1 being the lowest and 100 being the highest): ______ 2) Indicate the extent you ag ree this is a suitable can didate for the Senior Supervisor position. (Please circle the appropriate number) 3) Indicate the extent you agr ee that you would offer th is candidate the position of Senior Supervisor. (Pleas e circle the appropriate number) 4) Indicate the extent you agr ee that this candidate is your top choice for the Senior Supervisor position. (Pleas e circle the appr opriate number) Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Strongly Disagree Strongly Agree 1 2 3 4 5 6 7
58 Appendix C (continued) For Candidate B: 1) Please rate the applicantÂ’s overall cred entials based on the information given (1-100, 1 being the lowest and 100 being the highest): ______ 2) Indicate the extent you ag ree this is a suitable can didate for the Senior Supervisor position. (Please circle the appropriate number) 3) Indicate the extent you agr ee that you would offer th is candidate the position of Senior Supervisor. (Pleas e circle the appropriate number) 4) Indicate the extent you agr ee that this candidate is your top choice for the Senior Supervisor position. (Pleas e circle the appr opriate number) Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Strongly Disagree Strongly Agree 1 2 3 4 5 6 7
59 Appendix C (continued) For Candidate C: 1) Please rate the applicantÂ’s overall cred entials based on the information given (1-100, 1 being the lowest and 100 being the highest): ______ 2) Indicate the extent you ag ree this is a suitable can didate for the Senior Supervisor position. (Please circle the appropriate number) 3) Indicate the extent you agr ee that you would offer th is candidate the position of Senior Supervisor. (Pleas e circle the appropriate number) 4) Indicate the extent you agr ee that this candidate is your top choice for the Senior Supervisor position. (Pleas e circle the appr opriate number) Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Strongly Disagree Strongly Agree 1 2 3 4 5 6 7
60 Appendix C (continued) For Candidate D: 1) Please rate the applicantÂ’s overall cred entials based on the information given (1-100, 1 being the lowest and 100 being the highest): ______ 2) Indicate the extent you ag ree this is a suitable can didate for the Senior Supervisor position. (Please circle the appropriate number) 3) Indicate the extent you agr ee that you would offer th is candidate the position of Senior Supervisor. (Pleas e circle the appropriate number) 4) Indicate the extent you agr ee that this candidate is your top choice for the Senior Supervisor position. (Pleas e circle the appr opriate number) Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Please rank the applicants in the order that you would hire them. First: ____ Second: ____ Third: ____ Fourth: ____ Strongly Disagree Strongly Agree 1 2 3 4 5 6 7 Strongly Disagree Strongly Agree 1 2 3 4 5 6 7
61 Appendix C (continued) WITHOUT LOOKING BACK at the resumes, Please answer the following: Now we'd like you to consider the various crit eria that went into your decision. Please Rank Order the following criteria in order of importance for your decision. Place a "1" next to the criterion that you feel is the most important; place a "2" next to the second most important; and so on until you have ranked all 4 dimensions below: RANK Level of Education ___ GPA ___ Experience ___ Skills ___ Please circle the applicant with the highest GPA Candidate: A B C D Please circle the applicant with the most Experience Candidate: A B C D THANKS FOR YOUR PARTICIPATION!
62 Appendix D: Hypocrisy Manipulation Public Service Message This research project is concerned w ith how young people in contemporary U.S. perceive the problem of racial prejudice. We are interested in exploring techniques that might be effective in teaching young people about problems and solutions to racial prejudice. Specifically, we would like to create pub lic service messages using real college students. We are interested in finding e ffective essays that get across the message that racial prejudice and di scrimination are still problems that exist in our society. Excerpts from these essays may eventually be used in a public service pamphlet directed at high school students for the Â“Racial Equity Forum Â”. College students may be more credible with high school students because college students are seen as more experienced, but not so different that they would lose their credibility. The purpose of this study is to write a brief essay di scussing racial prejudice and discrimination. Your task will be to write a short essay on the following page highlighting the current problem of racial prejudice and discrimination. To assist in writing this essay, here is a list of facts from Â“The University of Michigan Documents CenterÂ” website: 1) The unemployment rate for Blacks has b een twice that of Whites for more than 20 years. 2) On average, Black males earn only 74% of White males with similar education. 3) On average, Hispanic males earn only 63% of White males with similar education. 4) The relative pay of college-educated Black men compared to collegeeducated White men has fallen by more than 10 percentage points in the last twenty years. 5) Between 1979 and 1997, the pay of Black women relative to that of White women fell by nearly 10 percentage points. 6) The average Black family income is about 40% less than the average White family income. This is the same as it was in 1967. You may use these facts above in your essa y if they help, or you may come up with your own message. Please notify the e xperimenter when you are finished.
63 Appendix D (continued) Public Service Message (control) This research project is concerned w ith how young people in contemporary U.S. perceive the problem of smoking. We are in terested in exploring techniques that might be effective in teaching young people about the dangers of smoking. Specifically, we would like to create pub lic service messages using real college students. We are interested in finding e ffective essays that get across the message that smoking is still a problem in our so ciety. Excerpts from these essays may eventually be used in a public service a nnouncement directed at high school students. College students may be more credible with high school students because college students are seen as more experienced, but not so different that th ey would lose their credibility. The purpose of th is study is to make a brief essay discussing the dangers of smoking. Your task will be to write a short essay on the following page highlighting the dangers of smoking. To assist in writing this essay, here is a list of facts from thetruth.com website. 1) Tobacco kills more people than AIDS, murder, suicide, fires, alcohol, and all illegal drugs combined. 2) If both a childÂ’s parents smoke, it is the equivalent of the child actively smoking between 60 and 150 cigarettes per year. 3) Tobacco companies know that 70% of smokers want to quit but canÂ’t. 4) Tobacco companies know that that of the smokers who try to quit only about 3% succeed. 5) Smokers are admitted to hospitals twice as often as nonsmokers. 6) The tobacco industry lets people believe that light cigarettes are better for you, when actually, they can be even worse. 7) Tobacco companies put ammonia in cigarettes, which makes your brain absorb more nicotine than it normally would. You may use these facts above in your essa y if they help, or you may come up with your own message. Please notify the e xperimenter when you are finished. Are you a regular smoker? Yes No
64 Appendix D (continued) Anagrams Part of Cognitive Psychology involv es the study of how people process information in short-term memory. Please unscrambl e the following words as fast as possible (you will only have a few minutes) and the answer the following question. If you cannot descr amble a word, leave it blank 1. dorw ________ 2. ehos ________ 3. alnp ________ 4. acef ________ 5. mepo ________ 6. einl ________ 7. rfou ________ 8. lfog ________ 9. glyu ________ 10. krow ________ 11. ilst ________ 12. hmeo ________ Please circle the 3 words that were the hardest for you to descramble.
65 Appendix E: PANAS CURRENT MOOD SCALE Emotion researchers are intere sted in the mood of college students at various times. This scale consists of a number of word s that describe diffe rent feelings and emotions. Read each item and then mark the appropriate answer in the space next to that word. Indicate to what extent you feel this way right now that is, at the present moment. Use the following scale to record your answers. 1 2 3 4 5 6 7 Not at all Extremely ___ interested ___ irritable ___ distressed ___ alert ___ excited ___ ashamed ___ upset ___ inspired ___ strong ___ nervous ___ guilty ___ determined ___ scared ___ attentive ___ hostile ___ jittery ___ enthusiastic ___ active ___ proud ___ afraid
66 Table 1. Correlation matrix for independent variables. MRS IMS EMS IAT MRS 1 IMS -.481** 1 EMS .263** -.051 1 IAT .031 -.048 .037 1 **. Correlation is signifi cant at the .01 level.
67 Table 2. Correlation matrix for dependent variables. Corrected Global Black Rating Corrected Global White Rating Likert Composite Black Rating Likert Composite White Rating Corrected Global Black Rating 1 Corrected Global White Rating .370** 1 Likert Composite Black Rating .603** -.024 1 Likert Composite White Rating -.036 .568** .082 1 **. Correlation is signifi cant at the .01 level.
68 Table 3. Cell means for (hypocrisy x applicant race) interaction. Black White Hypocrisy 82.7 77.1 Control 80.2 78.1
69 Figure 6. Distribution of IAT Scores. -1.5 -1 -0.5 0 0.5 1 1.5IAT Difference Score *Difference score is calcul ated by subtracting the aver age latency for stereotype congruent condition (White with Good) from the average latency for the stereotype incongruent (Black with Good). Higher Sc ores indicate bias towards White (longer responses when Black is paired with Good).
70 Figure 7. Mediation of hypocrisy. _____________________________________________________________________ ** p<.05 ** p<.01 Guilt, Shame Composite Hypocrisy Effect Difference between Black and White .116 .123** .131*