The main purpose of the fourth study was to test whether cognitive downward comparison boosts people's evaluations of their relationship in a positive direction. A second purpose of Study 1 was to explore to what extent self- and partner-enhancing social comparison may have a differential effect on relationship evaluation. It was assumed (see Chapter 3 of the present thesis) that people would consider self-enhancement more socially undesirable than partner-enhancement, which, in turn, may induce them to make public statements about themselves as a partner that are more modest than their statements about their intimate partners.

Following this line of reasoning, it was presumed that people will be more reluctant to engage in self-enhancing than in partner-enhancing downward comparison. Thus, it was supposed that the self-enhancement thought-generating task will be experienced as more difficult than the partner-enhancement thought-generating task, and as a consequence, that self-enhancement will exert a weaker impact on the evaluation of the relationship than partner-enhancement (e.g., Schwarz, Bless, Strack, Klumpp, Rittenauer-Schatka, & Simons, 1991; Wänke, Bless, & Biller, 1996).

It should be noted, however, that individual's self-reported evaluations may be biased, because, for instance, individuals may be unaware of their true sentiment (i.e., satisfaction) about the partner or the relationship, or because their answers to self-report items are influenced by the social desirability of response alternatives or by demand characteristics of the experimental situation. Therefore, in addition to dependent self-report measures, we employed a more indirect, unobtrusive measure of attitudes toward the relationship by recording the response latencies to the questions on relationship satisfaction and commitment.

Fincham, Garnier, Gano-Phillips, and Osborne (1995) provide an initial example of research incorporating the cognitive accessibility-based approach to attitudes (see Fazio, 1995) in the study of relationships. Fazio (1995) proposed that attitude accessibility reflects the strength of association in memory between a given object (e.g., one's partner or one's relationship with the partner) and a given evaluation (e.g., satisfactory), and that accessibility in memory determines the latencies of response to attitudinal questions.

We would like to suggest that cognitive downward comparison will not only enhance relationship satisfaction, but will also make the evaluation of one's relationship cognitively more accessible. Moreover, according to Fazio (1995), attitude extremity is typically related to the attitude's accessibility in memory, that is, to its salience. Assuming that participants would hold a more favorable attitude toward their relationship following the downward comparison thought-generating task than after the no-comparison thought-generating task, participants in the former condition were expected to show shorter response times to the questions of the dependent variables than those in the latter condition.