Clarifying the relationship between trait aggression and self-control using random item slope regression.

Abstract

Myriad psychological research evinces a negative association between self-control and aggression with some arguing for self-control failure as a cause of aggression. Recent literature suggests that the relationship between aggression and self-control is likely more complex and even positive in some cases. One source of such conflict in the literature could be the presence of unaccounted for random item slopes in commonly used measures of self-control which may inflate the likelihood of Type I errors. This study (N = 1386) tested the hypothesis that self-control would share random item slopes with the facets of trait aggression using random item slope regression. We measured trait aggression and self-control via two common self-reports: the Buss-Perry Questionnaire and the Brief Self-Control Scale. Our analyses revealed that the facets of trait aggression shared significant random item slopes with self-control and that many of these slopes were positive, rather than negative. We also found that Type I error inflation was evident in models that did not account for these random slopes. These findings may in part explain some of the conflicting results in the literature and that researchers interested in studying self-control and aggression should test for random item slopes.