Assessing Multiple Determinants of Relationship Commitment in Cohabiting Gay, Cohabiting Lesbian, Dating Heterosexual, and Married Heterosexual Couples

Abstract

In view of estimates that about two thirds of all first marriages are likely to end in separation or divorce (Martin & Bumpass, 1989) and evidence that the experience of dissolving a close relationship is among the most stressful life events experienced (Dohrenwend, Krasnoff, Askenasy, & Dohrenwend, 1978), professionals working with families are highly likely to encounter clients who are considering leaving a relationship. One of the reasons why the decision to leave a relationship is so fraught with difficulty is that commitment to the relationship --a psychological state that represents attachment to a partner and a desire to maintain the relationship for better or for worse (Rusbult & Buunk, 1993)--is multiply determined (Levinger, 1979; Rusbult, 1983; Sabatelli & Cecil-Pigo, 1985; Thibaut & Kelley, 1959). That is, many factors in concert influence the extent to which one is committed to a relationship. Although some attention has been directed at developing self-report measures of relationship commitment as a global psychological state (e.g., Stanley & Markman, 1992; Whitley, 1993), there appear to be no self-report measures that assess the multiple determinants of relationship commitment. From a clinical perspective, although it would be useful to know that one member of a couple coming to marital therapy reports low relationship commitment, it might be more useful to identify what specific determinants of relationship commitment are problematic. Accordingly, this article presents preliminary information regarding a measure that assesses multiple determinants of relationship commitment. The position that relationship commitment is multiply determined is derived from several researchers working within the interdependence theory tradition (Levinger, 1979; Rusbult, 1983; Rusbult & Buunk, 1393; Sabatelli & Cecil-Pigo, 1985; Thibaut & Kelley, 1959). An integration of these researchers' efforts leads to the postulate that a strong commitment to a relationship is independently influenced by six factors. These include perceptions that one derives many rewards from the relationship, that one incurs few costs to being in the relationship, that outcomes derived from the relationship meet some comparison level or internalized standard of what is expected, that the quality of alternatives to the current relationship is low, that investments made in the current relationship are high, and that barriers to leaving the current relationship are high. To the author's knowledge, no study to date has presented evidence (a) that these six determinants of relationship commitment are, indeed, distinct from each other and (b) that these six determinants of relationship commitment provide unique information regarding an independent assessment of relationship commitment. With regard to (b), Sabatelli and Cecil-Pigo (1985) showed that comparison level and barriers made unique contributions to an independent assessment of relationship commitment in a sample of married couples, and Rusbult (1983) showed that rewards, alternatives, and investments made unique contributions to an independent assessment of relationship commitment in a sample of college students in dating relationships. However, there appears to be no study in which all six determinants of relationship commitment have been studied. The present study reports data from a new self-report measure--The Multiple Determinants of Relationship Commitment Inventory (MDRCI)--in hopes of addressing some of the limitations in this area of study noted earlier. Three comments regarding the design of the MDRCI are in order. First, because relationship commitment is typically only one of several relationship-related constructs of interest in either research or clinical settings, the MDRCI was deliberately designed to be brief. In the current version of this measure, each of the six purported determinants of relationship commitment is assessed by four items. …