present study were not involved in the care of
their child with autism to any great extent when
the child had lower levels of behavior problems.
However, when the child had high levels of behavior
problems, it was more important and necessary
for the father to become involved. This difference
could explain the effects observed here.
The point that psychological factors may act
differently for mothers and fathers is not reduced
by the possibility that the effects identified in the
present research may be related closely to involvement
in child care. However, it is important not
to be seduced by gender differences that may be
related to a third variable. In the present context,
more research is needed, but the data could potentially
be more properly interpreted as suggesting
that self-efficacy may mediate the impact of
child behavior problems on mental health for
those parents involved closely in the care of their
child. However, self-efficacy may act as a moderator
for those typically less involved in child care.
Finally, the results of the present study may
also have implications for theory and practice in
relation to problem behaviors of children with developmental
disabilities. If we accept the hypothesis
that parental mental health is likely to be an
important determinant of the way in which parents
deal with behavior problems and interact
generally with their children, then a focus on selfefficacy
may have an indirect impact on the amelioration
of problem behaviors. Of course, interventions
designed to improve self-efficacy may
also have a direct impact on parents' behavior
(e.g., via skills-training interventions). These relationships
could be explored in future research.
These potential implications suggest that a theoretical
model of the impact of behavior problems
on parents should also refer to the effects of parent
behavior on children's behavior problems. A
model of this kind would assist researchers and
clinicians considerably.