In order to test the contention that violent comic books influence the processing of social information, a 2 (Comic Book) × 2 (Gender) multivariate analysis of covariance was conducted, with Trait Hostility, Like, and Humor as the covariates. Results indicated a significant multivariate main effect for Comic Book, F (3, 108) = 6.64, p < .01 and for the covariate, Trait Hostility, F (3, 108) = 5.87, p < .01. Follow-up univariate analyses of Violent Comics and Aggression 377 covariance (ANCOVAs) were then conducted for each composite variable. For Intent, the univariate main effect for Comic Book was significant, F (1, 108) = 11.8, p < .001.
Participants reading the extremely violent comic book suggested more retaliation than participants reading the mildly violent comic books. Results for Emotion indicated a significant Comic Book univariate main effect, F (1, 108) = 23.9, p < .001. Participants reading the extremely violent comic books inferred a more negative emotional state to the provocateur than did the participants reading the mildly violent comic books. No other significant differences were evident. Table 2 presents the estimated means and standard errors for the composite variables.
Consistent with previous research using overt provocation situations , the current findings support Bushman’s contention that an individual with
a chronically aggressive network will demonstrate biased social information processing. Furthermore, reading violent comic books appears to influence judgments related to intention, retaliation, and emotion in ambiguous provocation scenarios. These findings are consistent with research showing a link between aggressive themes in comic books and aggressive feelings and perceptions of social situations.