Abstract
This article explores the relations among differentiation of self, adult attachment, sexual communication, sexual satisfaction, and marital satisfaction, in a path analysis model. In a sample of 205 married adults, the path analysis results indicated that (a) differentiation of self had no direct effect on marital or sexual satisfaction, although it was significantly related to sexual communication; (b) adult attachment had a direct effect on marital satisfaction, but not on sexual satisfaction; (c) sexual communication is a mediating variable; (d) sexual communication was positively related to sexual satisfaction and marital satisfaction; and (e) no gender differences existed in the model.
Notes
1. Some of the most commonly used fit indices in the literature include the chi-square statistic, which indicates the magnitude of sample-model differences (ranges from zero to infinity); the comparative fit index, which is the proportion of sample variance/covariance explained by the model (ranges from 0 to 1.00); the Tucker-Lewis index, which is the relative fit of the path model against a “null model,” typically one in which all variables are assumed to be uncorrelated (ranges from 0 to 1). The root mean square error of approximation should be less than .05 and nonsignificant. In well-fitting models, the chi square statistic should be nonsignificant; comparative fit index and Tucker-Lewis index greater than .90 (Kline, 1991).