It was hypothesized that adult male inmates in Arizona who were convicted of nonviolent offenses would score significantly higher on academic achievement tests than those convicted of violent offenses. This hypothesis was tested in a descriptive, cross-sectional study. Data revealed no significant statistical difference between test scores of the two groups, thus negating the hypothesis. Based on the research results and a literature search, the author concludes that thinking errors are more significant than either intelligence or education in the choice to adopt a criminal lifestyle in general and in the incidence of violent criminal behavior in particular. He recommends increased efforts in the areas of critical thinking, morals education, and problem solving, either as separate courses or as components of existing programs. 16 references, 4 tables, and 2 graphs