Changing Racial Prejudice Through Diversity Education

Journal of College Student Development, Mar/Apr 2005 by Hogan, David E, Mallott, Michael

The Need for Cognition Scale (Cacioppo & Petty, 1982) consisted of 18 statements assessing how much the participants felt that the statement was characteristic of them (1: Extremely uncharacteristic to 5: Extremely characteristic). Sample items include: "I like to have the responsibility of handling a situation that requires a lot of thinking"; "I would rather do something that requires little thought than something that is sure to challenge my thinking abilities"; and "I really enjoy a task that involves coming up with new solutions to problems." Higher total scores on the scale reflect individuals who derive enjoyment from thinking and problem solving. Cronbach's alpha in the full sample was .91.

The Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1960) consisted of 33 statements to which the individuals responded true or false about their tendency to engage in socially desirable behaviors that are not commonly practiced and to engage in socially undesirable behaviors that are commonly practiced. Sample items include: "I never hesitate to go out of my way to help someone in trouble"; "I have never intensely disliked someone"; and "I like to gossip at times." Higher total scores on the scale reflect a relatively strong social desirability disposition. Cronbach's alpha in the full sample was .82.

RESULTS

Throughout the following presentation of results, calculated p values will be provided wherever possible. Due to data analysis software limitations exact p values could not be calculated when p was less than .001; therefore, those probabilities will be designated as p

Pretest Results

Two groups were formed on the Need for Cognition Scale (NFC) based on a median split. Participants in Group Low NFC (n = 56) were below the median score of 61, and participants in Group High NFC (n = 60) were at or above the median. The mean prejudice scores are presented in Table 2 as a function of Need for Cognition condition and Instruction Group. As anticipated, the students who measured high in need for cognition had lower prejudice scores than students who measured low in need for cognition, but the difference was small. Also consistent with expectations is the observation that Group R/G Com had the lowest average prejudice score, but the variation in means of the instruction groups was also very small. An Instruction Group (3) × NFC (2) ANOVA (see Table 3) revealed a nonsignificant effect of Instruction (p = .495), a marginally significant effect of Need for Cognition (p = .077), and a nonsignificant interaction (p = .429). Although the instruction and need for cognition effects did not reach statistical significance, one must keep in mind that the sample sizes of the pretested groups were relatively small so the absence of a significant difference could reflect low power (a Type II error).

Posttest Results

Mean prejudice scores on the postest, pooled across the Pretest/No Pretest subgroups, appear in Table 2. A Group (3) × Need for Cognition (2) × Pretest Condition (2) ANOVA revealed a significant difference among the Instruction Groups, F(2, 238) = 7.576, p = .001; a significant effect of Need for Cognition, F(1, 238) = 7.817, p = .006; and a nonsignificant effect of Pretest condition, F(1, 238) = 1.146, p = .285. No interaction effect in the analysis was significant (see Table 3). Pairwise comparisons of the instruction groups confirmed that Group R/G IP had significantly lower prejudice scores than both Group No R/G (p