MLK DAY AND RACIAL ATTITUDES 11
accessibility by conducting a two-sample t-test between the accessibility of MLK of individuals
who attended/planned to attend an event versus those who did not. People who attended/planned to
attend an MLK event indeed had higher accessibility (M=2.27, SD=1.22) of MLK than those who
did not (M=1.32, SD=.59), t(211)=7.63, p<.001, d= 1.15.
We conducted a 2 (Date: MLK Day versus not) X 2 (Holiday Accessibility: High versus
Low) MANOVA on Target Evaluation Level (Group Rating, versus Modern Racism, versus
Individual). There was a significant 3-way interaction between Target Evaluation Level, Date, and
Holiday accessibility, F(2,208)=15.02, p<.001 (see Figure 2). Decomposing this three-way
interaction, the 2-way interaction between Date and Target Evaluation Level was significant for
participants who planned to attend an MLK Day event, F(2,60)=11.84, p<.001, as expected.
Participants who attended or planned to attend an MLK Day event reported greater liking for
African-Americans as a group on MLK Day (M=6.50, SD=1.70) than otherwise (M=5.65,
SD=1.36), p =.03, d=.56. Consistent with Study 1, they also reported lower levels of modern racism
on MLK Day (M=1.50, SD=.43) than otherwise (M=1.87, SD=.69), p =.01, d=.66. Conversely, these
same participants rated Barack Obama more negatively on MLK Day (M=4.84, SD=2.27) than
otherwise (M=7.13, SD=2.51), p<.001, d=.97. As predicted, the two-way interaction was not
significant for participants who did not plan to attend an MLK Day event, F(2,147)=.51, p=.61.
Also as predicted, this interaction was not observed for ratings of the other control groups:
Republicans (p=.49), Democrats (p=.16), Independents (p=.48), Caucasian-Americans (p=.13),
Hispanic/Latino-Americans (p=.26), or Asian-Americans (p=.20). This effect was further not seen
for ratings of the other exemplars: white politicians (p=.92), white entertainers (p=.36), black
entertainers (p=.67), Herman Cain (p=.70), or Jesse Jackson (p=.42).
Finally, we conducted a binary logistic regression predicting event attendance (yes or no)
from age, gender, importance of MLK Day, political orientation, and race. Age was the only