To install an active, cognitive downward comparison process similar to that described by authors such as Taylor et al. (1983), Wills (1987) and Buunk and Ybema (1995), the current studies employed a manipulation of cognitive downward comparison that is new in social comparison research. Participants were induced to actively engage in cognitive downward social comparison by employing a thought-generating task in which participants were asked to explicitly describe positive qualities of their relationship which they consider to be better than most others (hereafter referred to as the comparison thought-generating task). Thus, participants are deliberately prompted to focus on the superiority of their own relationships over those of others, which, in turn, will supposedly boost individual's perceived quality of the relationship.

The present studies also included a control condition to test our assumption that it is specifically downward comparison that enhances the relationship evaluation, and not the fact that our instruction induced an increased salience of the positive features of the relationship. Participants in this control condition were presented with a similar thought-generating task but instead were asked to describe positive qualities of their relationship which they consider as good (hereafter referred to as the no-comparison thought-generating task). In contrast to participants assigned to the comparison thought-generating task, those in the no-comparison thought-generating task were instructed to make relationship-evaluations against an absolute, noncomparative standard. It was expected that in the former --downward comparison - condition individuals would subsequently evaluate their relationship more favorable than those in the latter --no-comparison-- condition.

The experimental paradigm was derived from Salancik (1974, 1976) and Seligman, Fazio and Zanna (1980). They reasoned that people's cognitions, attitudes, and beliefs can be affected by manipulating the recall of specific attributes and behaviors performed in the past, or by making specific dimensions more salient than others. Seligman et al. (1980) asked intimate partners to indicate either the intrinsic reasons or the extrinsic reasons for being in their relationship. The intrinsic reasons emphasized the couple's enjoyment of each other as motivations to maintain the relationship. Extrinsic reasons stressed the external rewards associated with their partner. It was found that the induced salience of extrinsic reasons led participants to express less love than those induced to the intrinsic cognitive set.

In each experiment participants took part in a study that was announced as a study intended to develop a questionnaire for which we needed individuals' thoughts about their intimate relationships. The experimental paradigm was a thought-generating task which asked participants to generate positive qualities on which they considered themselves (self-enhancement, Study 1), their intimate partners (partner-enhancement, Studies 2 and 3), or their own relationships (relationship-enhancement, Study 3) in noncomparative or comparative terms. Study 1 examined the effect of cognitive downward comparison on relationship quality. Studies 2 and 3 tested whether these effects were moderated by participants' relational discontent and their social comparison orientation.