Attributions, affect, and behavior in abuse-risk mothers: a laboratory study.

Objective: There were two main aims: first, to assess parental attributions about child behavior in abuse-risk and nonclinic parents. Second, to assess how attributions predict affective and behavioral reactions to child behavior.

Method: Internal-external attributions relating to the causes of child behavior were compared across mothers at-risk of child abuse (n = 40) and mothers who reported no significant parental or child conduct or behavior problems (n = 20). Mothers' attributions about the causes of the behavior of their own child and an unfamiliar child were recorded in response to the presentation of videotaped excerpts of the behavior.

Results: Results highlighted that compared with nonclinic mothers, abuse-risk mothers had a tendency to attribute positive child behavior to more external causes and negative child behavior to more internal causes. Differences were also found between parental cognitions about clearly positive, clearly naughty, and ambiguous child behavior. In the abuse-risk group, positive child behavior predicted coercive parenting when it elicited angry feelings in the mother; ambiguous and naughty child behavior led to coercive parenting through valence ratings of "deviant" and attributions of "internality." Analyses within the abuse-risk group showed that parental attributions are predictive of parental coerciveness for unfamiliar behavior. As behavior becomes more familiar, ratings of its valence and the affect it elicits override attributional activity.

Conclusions: Parental attributions about the causes of child behavior differ according to the valence and familiarity of that behavior, and discriminate between parents at risk for child abuse. Further, attributions are predictive of the affective and behavioral responses the parent makes to the child's behavior for ambiguous or unfamiliar behavior. Evidence was found for the validity of using videotaped stimuli of the behavior of known and unknown children as a method of assessing parental attributions.