Predicting aggressive behavior in children with the help of measures of implicit and explicit aggression

Abstract

Aggressive behavior between children in schools is a topic that receives much interest as violence and aggressive behavior cause many maladaptive social outcomes in the school setting. In the current study the Implicit Association Test (IAT) was adapted as a measure of children’s implicit aggression, by assessing the association of the self category (vs. other) with the attribute concept of aggressive (vs. peaceful). In addition to the IAT, a questionnaire measure to assess explicit aggression was utilized. The main goal of the present contribution was to examine the predictive validity of measures of implicit as well as explicit aggression in predicting the scores in a competitive computer game as an indicator of children’s aggressive behavior. Taken together, the results indicated that measures of implicit and explicit aggression could serve as reliable predictors of children’s aggressive behavior and that the IAT possesses incremental validity in addition to the self-report measure.

On the psychometric properties of the aggressiveness‐IAT for children and adolescents

“…Van Goethem et al (2010) report test-half reliabilities between r ¼ .72 and r ¼ .74 for two different variants (i.e., one with unprimed and one with primed bullyingrelated stimuli) in a sample of primary school children. Grumm et al (2011) report an adjusted split-half correlation of r ¼ .74 for this age group. Internal consistencies (Cronbach's a) varied between a ¼ .55 (Bluemke & Zumbach, 2007), a ¼ .81 (Teubel et al, 2011), and .89 a .91 (Richetin et al, 2010).…”

“…Notably, researchers have employed the Agg-IAT to capture both stable individual differences in aggressiveness (Banse & Fischer, 2002;Gray, MacCulloch, Smith, Morris, & Snowden, 2003;Grumm et al, 2011;Richetin et al, 2010;Teubel et al, 2011;Van Goethem et al, 2010) and changes due to manipulations or interventions (Bluemke & Zumbach, 2007Uhlmann & Swanson, 2004;Gollwitzer et al, 2007;Bluemke et al, 2010). In other words, some studies have used the Agg-IAT on the (implicit) assumption that it measures a trait that can predict aggressive behavior across situations and time, whereas other studies have used it on the (equally implicit) assumption that it can reliably capture true fluctuations between situations.…”

Attitudes Towards the Use of Violence and Partner-Directed Aggression

“…Implicit bullying attitudes measured with an IAT have been shown to act as moderators in predicting school bullying behavior for children with relatively positive explicit bullying attitudes (van Goethem, Scholte, & Wiers, 2010). An aggression IAT was shown to predict aggressive behavior in children over and above the variance accounted for by explicit self-report measures (Grumm, Hein, & Fingerle, 2011). Another aggression IAT indicated that the more closely one implicitly associates themselves with the attribute of aggressive, the more aggressively they respond to provocation (Richetin, Richardson, & Mason, 2010).…”

Alternatives to Self‐Reports

“…Thus, quicker sorting of stimuli related to self and aggressive with one response key and other and peaceful with the other response key (block 3), contrasted to the reversed pairing (block 5), will result in a positive IAT-A score. More importantly, it has been shown that the IAT-A is mostly unrelated to aggression and predicts specific aggressive behaviors over and above the corresponding direct questionnaire measures (incremental validities ranged from 5-15% of variance in aggressive behavior: Banse et al, 2015;Grumm et al, 2011;Teubel et al, 2011). Previous studies have shown satisfactory reliabilities of the IAT-A (Cronbach's Alpha ranged from 0.69 to 0.91; Banse et al, 2015;Richetin, Richardson, & Mason, 2010;Teubel, Asendorpf, Banse, & Schnabel, 2011;Spearman-Brown split-half 0.74, Grumm, Hein, & Fingerle, 2011).…”