Running on empty

This is the authors' abstract. We don't add key points for this paper.

Commitment has often been used to examine why individuals unjustifiably persist in relationships that are lacking in or devoid of satisfaction. However, the practicality of using commitment to examine these situations has been questioned because of its substantial association with satisfaction. Across three studies, we created a measure of nonvoluntary dependence and investigated the validity and reliability of the Nonvoluntary Dependence Scale from an investment model perspective. Exploratory and confirmatory factor analyses revealed evidence of a single factor and reliability analyses revealed good internal reliability for the measure. The new measure also evidenced desirable convergent and discriminant validity with respect to a number of existing individual- and relationship-level constructs. Consistent with hypotheses, nonvoluntary dependence was significantly associated with commitment level, investments, and alternatives but not with relationship satisfaction. Differences in nonvoluntary dependence were also found between individuals in aggressive versus nonaggressive relationships, with no differences found in commitment level. Implications regarding this new measure, as well as directions for future research, are discussed.