Approximately 2 months before scheduling the experimental sessions in the laboratory, participants were recruited via electronic mail, which asked for undergraduates to participate in a study about relationships. The cover story was essentially the same as in the fourth study. It was mentioned that interested participants were eligible to participate in the study if they were involved in an intimate relationship for at least the previous 3 months.

Two hundred and eight potential participants responded via electronic mail. Potential participants were sent questionnaires by mail, together with a postage-paid return envelope and a cover letter that thanked them for participation in the research. They were asked to complete the questionnaire individually and to answer all questions sequentially. A reminder letter was sent to those participants who did not return the questionnaires within two weeks.

Following this procedure, 166 participants returned completed questionnaires, yielding a response rate of 78%. They were asked to supply their names and telephone numbers so that they could be contacted at a later date. A financial incentive of 5 Dutch Guilders (approximately U.S. $3) was offered if they completed and returned the questionnaire.

Approximately 3 to 5 weeks after returning the questionnaires, one hundred and eighteen participants could be contacted by telephone and a time was arranged for them to come to the laboratory. One hundred and twelve participants showed up for their scheduled follow-up session in the laboratory. Two participants reported that they were no longer involved in relationships with their original partners and were dropped from the sample.

Therefore, the final sample consisted of 110 University of Groningen undergraduates (74 women, and 36 men). Ninety-four were involved in dating relationships, 14 participants were cohabiting, and 2 were married. The average duration of their relationship was 2.0 years, with a range of 3 months to 8.4 years. Participants were 21.8 years old on average (range 18 to 29 years). All participants were paid 7.5 Dutch guilders (approximately U.S. $4.5) for their participation in the laboratory research.

The questionnaire was designed to assess social comparison orientation, initial levels of relational discontent, and prior levels of relationship satisfaction and commitment to the relationship. The INCOM (Gibbons & Buunk, 1999) was used to measure social comparison orientation (Cronbach's alpha = .83).

Subsequently, a 7-item self-report scale was constructed to assess relational discontent. The items were found to load on a single factor and the reliability of the composite was high (Cronbach's alpha = .82). Participants indicated, on 9-point scales, the extent to which they were discontent with their relationship, as well as the extent to which they had been bothered or distressed by the course of events in their relationships during the past two months. The scale included, among others, the single item used in Study 2 to assess discontent with their partner (1 = completely content and 9 = totally discontent). In addition, the scale included items such as "Have you or your partner seriously discussed the idea of ending the relationship during the past weeks?" (1 = not at all and 9 = very frequently), "Do you feel that your relationship is under a lot of pressure due to unresolved differences of opinion between you and your partner?" (1 = no pressure at all and 9 = a lot of pressure), and "Do you feel that you don't know were you stand in your relationship with your partner?" (1 = completely disagree and 9 = completely agree).

In addition, prior levels of the dependent variables were assessed. Thus participants completed measures of satisfaction (Cronbach's alpha = .92) and commitment (Cronbach's alpha = .85) that were identical to those used in Studies 1 and 2, and were asked to evaluate the overall quality of their relationship by placing a mark on a 100-millimeters line, anchored at zero with "extremely low quality," and at 100 with "extremely high quality." The distance measured in millimeters from the left end of the line was assessed to measure perceived quality of the relationship, resulting in a nearly continuous measure.

In the laboratory, the procedure was identical to Study 1, with the exception that, instead of manipulating self- and partner-enhancing downward comparison, participants in this experiment were instructed to focus on the qualities of their relationship which they considered as good or as superior to other relationships. The experimenters were unaware of participants' relational discontent levels and comparison orientation throughout the experimental procedure, and assignment to the no-comparison (n = 55) versus downward comparison condition (n = 55) was randomly determined.

On each table, the experimenter had placed an envelope. The participants were told that the envelope contained test materials and that they had to start with the computer-administered procedure by pressing a key, following the instructions. They were asked to work their way through the program at their own pace and to refer to the envelope when they were requested, to complete the questionnaire it contained. Each individual was then led to an individual booth and seated in front of a microcomputer.

After completing the thought-generating task, participants were presented with the dependent measures, that is the measures of satisfaction and commitment that were the same as in the pretest (Cronbach's alphas = .95 and .87, respectively), including the 100-millimeters measure of perceived quality of the relationship that was administered as a paper-and-pencil question. The questions of satisfaction and commitment were answered by pressing a numerical key (1 through 9) on the keyboard of the computer. Next, participants were instructed to open the envelope and complete the paper-and-pencil question. The response latencies to the questionnaires of satisfaction and commitment, as well as baseline response times, were recorded following the same procedure as in Study 1.