This study tested Rusbult's investment model of relationship commitment and stability using data from both partners of 167 heterosexual couples studied from 1972 to 1987. Multiple regression analyses generally confirmed predictions that rewards and costs account for a significant proportion of the variance in satisfaction and that satisfaction, investments, and quality of alternatives together account for a significant proportion of the variance in commitment. Evidence about the hypothesized mediational role of subjective commitment in predicting the duration of relationships was mixed. Path analyses showed that Rusbult's model provided an adequate fit to the data and that the associations among variables were similar for men and women. The model successfully predicted relationship duration over a 15-year period. Limitations of the model and directions for future research are considered.

No abstract