In general, Study 3 was highly similar to Studies 1 and 2, except for a few modifications. First, in an effort to reduce possible carryover effects the pretest of relational discontent and comparison orientation were administered approximately four weeks prior to the experimental part of the research and the measurement of the dependent variables. Second, the pretest measure of discontent was expanded to include some additional items measuring discontent with the relationship. Third, as the two previous studies could not test the assumption that cognitive downward comparisons enhanced perceived relationship quality, Study 3 examined if satisfaction and commitment did indeed increase due to cognitive downward comparison by including pretest measures of satisfaction and commitment. Fourth, in addition to the measures of general relationship satisfaction and commitment to the relationship, an additional dependent variable was employed that asked participants to evaluate the overall quality of their relationship by placing a mark on a graphic line.

We assumed this to be a more sensitive measure of perceived relationship quality than the general measures of relationship satisfaction and commitment. Fifth, as in Study 1, response latencies as an unobtrusive measure of relational satisfaction were also assessed. Sixth, because we wanted to examine the effects of cognitive downward comparison on the global evaluation of the relationship, we examined the effect of relationship-enhancement instead of partner-enhancement on perceived relationship quality. The thought-generating task thus manipulated relationship-enhancing downward comparisons by asking participants to generate reasons why they considered their own relationship to be a good relationship or a better relationship than that of most others. In conclusion, the predictions were identical to those in Study 2.