Features of males on correctional supervision related to criminal behavior: Associations between executive functions, locus of control, perceived parental expectations, and severity of crimes committed

Criminal behavior is a problem of considerable magnitude in the United States. The purpose of this quasi-experimental study was a tripartite examination of the role of executive functions, locus of control, and perceived parental expectations concerning criminal behavior in males. The sample consisted of 122 males on correctional supervision along with 59 undergraduate college students who functioned as the comparison group. It was hypothesized that impaired executive functions, an external locus of control, and low perceived parental expectations would be contributing factors to criminal behavior, specifically concerning severity of crimes committed. The Behavior Rating Inventory of Executive Functions- Adult Version (BRIEF-A) was used to assess executive functions while the Adult Nowicki-Strickland Internal/External Locus of Control Scale (NS) was used to measure locus of control. The Shipley Institute of Living Scale (SILS) was utilized to measure overall intellectual ability and a new instrument was created by the author to measure perceived parental expectations. Multiple regression results revealed that seriousness of criminal behavior could be predicated, although with only 14.2% of the variance accounted for. Three predictors of criminal behavior included the BRIEF-A Behavior Regulation Index score, the locus of control score and the perceived parental expectations score. The hypothesis of impaired executive functions contributing to criminal behavior was supported with the BRIEF-A BRI having a significance level of p = .001. The hypothesis of external locus of control contributing to criminal behavior was supported with the LOC having a significance level of p = .048. The hypothesis of low perceived parental expectations regarding prosocial behavior and academic achievement contributing to criminal behavior was statistically significant with p = .029. Logistic regression results revealed the model proposed in this study correctly predicted group membership (more serious versus less serious criminal behavior) 84% of the time.