One hundred and sixteen participants who were at the time of the study involved in a relationship (74 married, 24 cohabiting, 18 dating exclusively) participated in the study. The sample included 61 women, and 55 men. The average duration of their relationship was 15.9 years (range 5 months to 52.0 years). Participants were 41.3 years old on average (range 18 to 78 years). All participants received a box of chocolates for their participation.

To obtain a wide range of participants, we recruited a sample of train travellers. As soon as the intercity train left the railway station, a researcher approached potential participants and requested their cooperation in a study on intimate relationships. The cover story was essentially the same as in the fourth study. Participants were asked if they were currently involved in an intimate relationship, and, if so, they were asked to participate in the research. Participants were provided with a booklet containing all measures and the experimental thought-listing task. They were randomly assigned to conditions in which positive qualities of the partner were generated in either noncomparative or in comparative terms.

After providing general demographic information about gender, age, and duration of the relationship, participants answered questions on social comparison orientation and discontent with their partner.

Social Comparison Orientation was measured by the Iowa Netherlands Comparison Orientation Measure (INCOM). This 11-item scale was developed by Gibbons and Buunk (1999) on the basis of a larger item pool, and was tested in more than 25 studies. In Chapter 1 evidence for the predictive validity of the scale was presented, and the most important correlates of the scale were described. In addition, evidence for the construct validity of the scale is provided by the high correlation (r = .66) between this scale and the Attention to Social Comparison Information Scale (Bearden & Rose, 1990). The INCOM does not correlate with social desirability, has in general a reliability > .80, and has a test-retest reliability over eight months of .72.

In an attempt to overcome participant's reluctance in reporting their interest in social comparison information, as outlined above, the instructions for these items in this scale read:

The measure consists of statements reflecting individual's social comparison activities and interests. Sample items are: "I often compare myself with others with respect to what I have achieved in life", "If I want to learn more about something, I try to find out what others think about it", "I always pay a lot of attention to how I do things compared with how others do things", and "I am not the type of person who compares often with others" (recoded). The items were measured using 5-point scales, ranging from strongly disagree (1) to strongly agree (5). All ratings were averaged to form a single index, resulting in a INCOM-score with the higher score indicating stronger social comparison orientation (Cronbach's alpha = .77).

Next, participants were asked to consider their feelings of discontentment about the relationship with their partner and to ignore their sexual dissatisfaction. More precisely, participants were asked "Does your life with your partner bring you contentment (putting aside the sexual side)?" Participants were asked to rate the extent to which they were discontent with the relationship with their partner on a 9-point scale (1 = completely content and 9 = totally discontent). Higher scores thus indicated higher levels of discontent.

Next, the experimental part of the booklet consisted of the manipulation of cognitive downward comparison. Most aspects of the procedure were similar to the fourth study. Half of the participants were asked to list all the reasons they could think of why they had good partners (no-comparison condition, n = 60), and half were presented with the question to describe why they had better partners than most others (downward comparison condition, n = 56). All participants were instructed to generate at least five reasons of the specified type they could think of. The rest of the page was left blank except for the lines on which they were asked to write down each new reason. As Study 1 was not a computer-administered experiment, participants could not be instructed to enter the five most important reasons into the computer. Therefore, the next page of the booklet contained the instruction to arrange the listed reasons according to importance, by adding a number to each reason (a 1 for the most important reason, and so on). Participants were instructed to write each number down the left margin.

Subsequently, participants responded to questionnaires that tapped the dependent variables. The measures of satisfaction (Cronbach's alpha = .93) and commitment (Cronbach's alpha = .75) were identical to those used in Study 1. Upon completion of the questionnaire, participants were thanked for their time. Instead of debriefing immediately, participants were provided with a postage free response card to request further information regarding the research. On receipt of a response card, a summary of the research findings was forwarded.