may recall expectations more in line with what is actually occurring with their division of

labor; the actual level of task differentiation, as reported by the wives, was still an

important predictor (at least for child-related tasks) but was secondary to how satisfied or

dissatisfied the wives were with their current roles. These two variables, role

dissatisfaction and child-related task differentiation, accounted for 32% of the variance in

wives' violated expectations scores.

 For husbands, individual well-being was identified in the analysis as a significant

predictor of their violated expectations score. Husbands' higher levels of individual well-

being were associated with violated expectations scores indicating the partners' actual

involvement met the expectations husbands had formed while their wives were still

pregnant, or that the husbands were more involved in housework and child care than they

had expected to be. Previous research similarly found when men contributed less to

housework and child care than they had expected, they reported higher rates of

depression (Strazdins et al., 1997). It may be husbands who have a more positive sense of

well-being are better able to be involved in housework and child care. It may also be that

husbands who meet or exceed their own expectations for involvement in these roles

experience a heightened sense of positive well-being as a result, while those who do not

meet or surpass their expectations for involvement go on to experience more feelings of

anxiety and depression. This would fit with Hawkins and Roberts' (1992) suggestion that

men may experience cognitive dissonance if they do not measure up to the new ideology

of involved fatherhood.

 A second predictor of husbands' violated expectations scores was their rating of

how well the volatile couple conflict type represented the couple's usual conflict style.