A Test of Competing Cognitive Explanations for the Boomerang Effect in Response to the Deliberate Disruption of Media-Induced Aggression

Abstract

This study advances research on the boomerang effect in response to anti-aggression media literacy interventions. Previous findings indicate that elementary school children can become more aggressive after exposure to such interventions. We test two competing explanations for the boomerang effect, media priming and psychological reactance, in a 2 × 2 factorial experiment with elementary school children (N = 128). Findings indicate that children may cognitively process antisocial elements of an intervention program in a manner that runs counter to the intended effect of prosocial messages. Specifically, children who were exposed to a media literacy intervention with violent media clips as examples reported an increase in willingness to use aggression, whereas children who were exposed to the same lesson without the clips did not. Therefore, the boomerang effect is best explained by the processing of violent clips (media priming) and is not likely due to resistance to the instructional elements of the lesson (psychological reactance). Implications for researchers and practitioners are discussed.

Notes

1. The ANOVA depicted in this illustration on raw change scores introduces error into the analysis due to ceiling, floor, and regression to the mean effects (Bereiter, 1963, Cronbach & Furby, 1970, Raykov, 1993). This correlation between Time 1 (where there is no difference between experimental groups on aggression) and the change score reduces the probability of finding the true effect because it is not independent of Time 1. Thus, in Figure 1, our p is .07 under a repeated measure ANOVA on more parsimonious raw aggression scores as the dependent variable, presented for visual purposes. The analysis, using the less parsimonious residual change scores, yields the same statistical result as an ANOVA controlling for Time 1, which we reported above.