71
average of 18 critical/sarcastic statements from comparison fathers. Inconsistent with previous findings, total commands used during the 30 minutes were not significantly different between the two groups of fathers. Eyberg and Robinson (1982) found that fathers of clinic-referred children issued more commands than fathers of normal children. Their analysis, however, was based on a shorter observation period (i.e., 10 minutes) that did not include a clean-up situation. Bessmer (1996) found that mothers of clinic-referred children used significantly more commands than mothers of non-referred children, and that a higher percentage of those commands were direct.
Although untreated clinic-referred mothers and fathers tend to use predominantly direct commands (72% in Bessmer's study and 70% in this study), their children are not more compliant with their requests. Direct commands alone do not appear to lead to child compliance. During PCIT, the parent is taught to utilize direct commands so that it is clear to the child what behavior is expected. Yet it may be other aspects of their commands (e.g., timing, use of commands which give the child opportunity to comply), other behaviors (consistent consequences for compliance and noncompliance), or the improved relationship overall that account for increased compliance. The ratio of direct commands, then, may not change from pre- to post-treatment, but other aspects of parents' behavior do change to reduce the child's noncompliance.
Other important differences were noted in the paternal behaviors between the clinic-referred and non-referred groups. Consistent with Bessmer's (1996) findings for mothers, parent prosocial behaviors differentiated the two groups of fathers. The summary variable of father Prosocial Behavior included the categories of