76
possible that a comparison of sequences of behaviors would differentiate referred and nonreferred parent-child dyads. For example, future studies could evaluate the conditional probabilities of behaviors following child and parent prosocial behaviors, such as praise, behavior descriptions, and reflections, as well as inappropriate behaviors, such as noncompliance, criticism, and whining.
In summary, the present study demonstrates that DPICS II can be used reliably and is valid for use with children with behavior problems and their fathers. The summary variables measuring child compliance and the inappropriate behavior of children appear to be most useful in distinguishing between clinic-referred and non-referred father-child dyads. The DPICS II fulfills the criteria established by Roberts and Forehand (1978) as an assessment device for children and has a variety of applications both in clinical and research settings. The present study clearly supports the inclusion of fathers in observational studies of children with disruptive behavior.