The study of feelings poses some special methodological problems because of these gaps between feelings and final report. The most obvious problems arise because the only way we can tell what someone is feeling is to ask him. The problem with self-reports is, of course, that the person may not report his feelings accurately. In all the self-perception research programs to be discussed in later chapters, the experimenters were very much aware of the danger that participants would be too cooperative and so took various measures to minimize the possibility. Since most of these defensive procedures are specific to the particular research programs, they are best described in the later chapters. However, I would not want to leave any doubt about this issue as you are reading about the research, so I will quickly summarize the kinds of procedures that have been employed.
First of all, the participants were often elaborately deceived about the actual purposes of the research, so that they would not know what they were expected to do. Whenever possible, the experimenter was left in the dark as well, at least about whether a particular participant was expected to do one thing or another. In addition, usually participants were carefully interviewed after the experiment was over to discover whether they had seen through the deception.
In other experiments the manipulations were so cleverly disguised that no participant could perceive their purpose. In some kinds of research the critical variables were autonomic responses such as heart rate and skin conductance that even the most eager-to-please participants could not deliberately control. Finally, many relationships were predicted, and observed, between self-reports of feeling and other variables that participants could not have known their position on, such as whether they were field dependent or not, or could not have changed, such as whether they were overweight. In studies of this sort, some participants were expected to respond in one way, others in a different way, and the experimenter was blind as to which kind of response a particular participant should make. Thus, no consistent pattern of experimenter bias and participant compliance could have produced the observed results. Furthermore, in many of these studies, "faking" correctly would have required the participants themselves to discern very specific and anti-intuitive theoretical predictions about how they should behave. In sum, and as will become apparent as I discuss the actual research, few if any of these results can be explained away as due to participant compliance.
Some perspective on the risks posed by eager participants comes with the recognition that this problem is not unique to research on emotions, motives, or attitudes. For example, when a perception researcher shows participants the Muller-Lyer illusion and asks which line seems longer, the researcher must depend on the truthfulness of the participants' reports.
In fact, self-reports are like any measure of an entity that is not directly observable by the scholar's senses. Confidence in any measure of a complex, abstract entity develops through construct validation. That is, we make theoretically guided predictions about how the underlying entity and its measurement would relate to other measures or events in the world. To the extent that our measures behave in the way that the theory predicted, we gain some confidence in each component, the theory that made the prediction, the theoretical construct that was part of this theory, and the practical measurement of this construct. For example, self-perception theory predicted that women who were more responsive to personal cues would report feeling lowered self-confidence while reading women's magazines (Wilcox & Laird, 2000). When we found that after reading the magazine, the predicted group of women responded to a self-esteem scale with lower scores, we felt more confident about our theory, and also about our measurements of personal and situational cue use and of self-esteem. One such success conveys only a small bit of confidence in the measures, but repeated successes justify increasing confidence.
Certainly, in the case of self-reports of feelings, the construct validation work has been done long ago. Literally thousands of experiments have used self-report measures of feelings and observed the expected relationships. In fact, few, if any, psychological constructs and measures have accumulated greater stores of construct-validational confidence.
Of course, occasionally self-report measurement of feelings may err. A participant may lie to please or confound the experimenter, just as occasionally an intelligence test may fail to identify a good student, or an ohm meter may not measure electrical resistance correctly. It is a hallmark of an adequately validated measure/theory complex that it contains, at least implicitly, the criteria for identifying the occasional exception to its usual good functioning.
Although we have strong theoretical and empirical grounds for relying on self-reports about feelings, we should recognize that self-reports are, of course, distinct from the feelings reported. As noted earlier, there is clearly more to experience than we talk about, and perhaps ever could talk about.
For example, when I look at a painting, I could say that I feel peaceful and enraptured by what I see. My words do not begin to capture all that I am feeling, just as when I say that the picture depicts a three-dimensional landscape, those words do not capture all the experiences that come from the juxtaposition of objects in the three-dimensional "space" of the picture. Experiences of physical space and of feelings are complex and multidimensional (e.g., deRivera, 1977; Ortony, Clore, & Collins, 1988). How much of that complexity we try to capture in verbal reports is, of course, a matter of what our theoretical purpose is.
Was this article helpful?
I already know two things about you. You are an intelligent person who has a weighty problem. I know that you are intelligent because you are seeking help to solve your problem and that is always the second step to solving a problem. The first one is acknowledging that there is, in fact, a problem that needs to be solved.