The statistical results concur closely with descriptive ones presented earlier, indicating that the latter are not results of random variations. The main effects of employment, the contingent effects of parenthood, and the time trends modeled here are the same as reported earlier. (Only two differences occur. First, among white married women, mothers have statistically more acute conditions than nonmothers. We saw this parenthood effect earlier for nonemployed married women, but not for employed ones [results were inconsistent]. The statistical analysis smooths those inconsistencies and reveals that employed mothers also experience more acute conditions than their nonmother peers. Second, among white married women, older housewives show statistically increased chronic limitation over time. Earlier we saw a rise for housewives without children. The latter parenthood effect is statistically smaller than the age effect.) The singular advantage of the statistical analysis has been its ability to highlight interaction effects among the variables, some of which were not considered in the descriptive section. Comparing the models, note how those for short- and long-term disability are very similar to each other but distinctly different from the acute-condition models. This means that social roles and age influence short- and long-term disability in the same way. Specifically, both are greater for older and nonemployed women, being especially high for older nonemployed women and housewives without children. By contrast, the most consistent factor affecting acute-condition incidence and impact is presence of children. Children increase their mothers' experience of acute problems but reduce the amount of recuperative time and medical care taken for them.
Read full abstract