How are questionnaire data similar to, and different from, thought-sampling data? Five studies manipulating retrospectiveness, single-moment focus, and indeterminacy

Abstract

Results reported in the literature from thinking research are difficult to compare because the various studies differ on three independent dimensions: whether the observations are immediate or retrospective, whether the observations are targeted at measuring single moments in time or multiple moments, and whether the observations are directed at specific thought events or at characterizing thought events in general. Furthermore, questionnaire and sampling studies typically collect data on differing sets of variables. In an attempt to separate the influences of these dimensions in the thoughtsampling literature, factor-analytic techniques were applied to five studies in which subjects described their thoughts and moods on the same 42-variable rating questionnaire. Studies were designed to differentiate between factors produced using immediate versus retrospective techniques, single-moment versus multiple-moment techniques, and specific-event versus indeterminate-event techniques. Factors that had occurred in previous research called Aggressive/Bad Mood, Pleasant/Sexual, Clear Thought, Daydream/Past Sexual, Duration, and Self-Critical were found in all five studies. However, factor-loading patterns in the five studies were somewhat different, leading to the conclusions that the retrospective nature of questionnaires is likely to distort the frequency of recollections in thinking; that the indeterminate nature of retrospective questionnaires might lead to the erroneous conclusions that clarity in thinking is associated with pleasant thoughts and to a minimization in the reports of strength of negative affect; and that either the multiple-moment nature or the between-people analysis of retrospective questionnaires might lead to exaggeration of the relationship between sex and aggression in experience.