Longitudinal Data Relating Marriage Satisfaction and Having a Child

LeMasters (1957) reported that having a child frequently led to a family crisis. Unfortunately, his data consisted of retrospective self reports. There was no control over constant error in the judgments made, i.e., there was no comparison group that had not had a child. Also, little information was provided regarding the modal kinds of reported feelings which were characterized as a crisis. Subsequent studies also using retrospective reports and no comparison group, but with varying sample composition, varying retrospective periods, and varying measurement procedures yielding widely varying percentages of crisis cases (Dyer, 1963; Hobbs, 1965). The work of Feldman and his associates added some clarity in this area, in that actual longitudinal data (Feldman, 1971) indicated a postpartum decrement in scores of the sort usually labeled marriage satisfaction. His subjects reported more negative things about their marriages as a correlate of having had a child. Feldman and Rogoff (1968) went further, and suggested, in effect, that this downward shift interacted with how contented the couple claimed to be prepartum. Apparently, couples who were more contented prepartum were more affected by the birth of a child than were couples who were less contented in the prepartum period. The Feldman and Rogoff data, however, provide no way to distinguish between a genuine interaction effect and regression to the mean. To really demonstrate the kind of effect suggested by Feldman and Rogoff requires a comparison group, which they did not have.