A pilot study using a computer-based rule following task to distinguish adolescents with and without a behavior disorder

The purposes of this pilot study were to: (a) determine if adolescents with a behavior disorder were less likely than adolescents without a behavior disorder to follow rules when the contingencies did not support doing so, and (b) explore the possibility that a computer-based rule-following task could be used as a measure to differentiate children with and without behavior disorders. Thirteen adolescents with a diagnosis of a behavior disorder (Conduct Disorder or Oppositional Defiant Disorder) and thirteen adolescents without the diagnosis of a behavior disorder were exposed to a computer task in which the rule for earning points and the contingency for earning points did not match. A comparison between groups using a Mann–Whitney U analysis revealed that students with an EBD were significantly more likely to violate the rule, z = −3.724, p < .01. A discriminant analysis yielded a significant Wilks’ Lambda (λ = .490, p < .0001), canonical correlation (Rc2=.714, p < .0001), and a correct classification rate of 82.7%. Implications of these findings are discussed in terms of measuring rule-following and using the computer-based task as a screening tool to identify students who are in need of further assessment for behavior disorders.

► Adolescents with and without behavior were compared on a rule-following task. ► Adolescents with a behavior disorder violated the rule more often than those without. ► The computer-based task proved to be a reliable measure of rule-following behavior. ► The task was accurate in predicting which adolescents were diagnosed and those who were not.