Key points are not available for this paper at this time.
In this paper we note the inconsistent findings that characterize research on marital instability and propose several potential methodological sources of these inconsistencies. The issues addressed range from problems of external validity introduced by limited samples to the constraints imposed by the various multivariate analytic techniques utilized in previous research. An illustrative application to data from two recent nationally representative probability samples highlights the methodological pitfalls that characterize this area of inquiry. It is argued that, unless more attention is devoted in the literature to issues of method and measurement, little consensus on the determinants of marital disruption can be expected.
Hanson et al. (Fri,) studied this question.