Spiers, Cindy J. 「Commitment and Stability in Lesbian Relationships.」 Ph. D diss., University of Maryland College Park, 1998.

Abstract:
This study of 227 self-identified lesbians, recruited primarily through the Internet, was designed to apply a model of relationship stability developed with a heterosexual population to a lesbian population. Specifically, the model tested was based on Stanley and Markman's (1992) research using the Commitment Inventory, which breaks commitment into two factors: personal dedication, such as a strong sense of 'couple identity', and a belief in the importance of commitments, and external constraints to leaving a relationship, such as financial and legal bonds. Stability was operationalized in two ways: a self-report measure comprised of 5 items used by Rusbult (1983), 3 items used by Peplau, Padesky and Hamilton (1982), and 7 items from Sternberg's (1988) triangular assessment of love scale, and actual relationship longevity. Three other variables, self-disclosure of sexual orientation, self-acceptance of sexual orientation, and satisfaction were included in the model as moderator variables in the hypothesized relationship between personal dedication, external constraints on a relationship, and stability. Regression analyses found two different models were supported depending on the measure of stability used. Specifically, 59% of the total variance in self-report stability was accounted for by personal dedication, satisfaction, self-disclosure, the interaction between personal dedication and self-disclosure, and the interaction between satisfaction and constraints on a relationship. Of the variance accounted for by this model, three-fourths was explained by personal dedication alone. In contrast, only 3% of the variance in actual relationship longevity was accounted for, all of which was explained by constraints on a relationship. The correlation between the two measures of stability was.21 ($p<.01).$ These findings closely parallel findings from research on heterosexual couples (Adams and Jones, 1997; and Stanley and Markman, 1992) Results seem to indicate that aspects of commitment based on self-perception are predictive of one's perception of relationship stability, whereas more behavioral aspects of commitment are predictive of actual staying together in relationships. Recently, Adams and Jones (1997) have argued for refining the definition of commitment to include only intentionality aspects, such as personal dedication, with factors such as constraints being understood as a separate construct which act in concert with commitment to affect the endurance of a relationship. The findings from the current study support this conclusion.