The term introspection is generally used by psychologists to refer to people’s observation and contemplation of their own thoughts, feelings, and sensations. In early psychology, trained introspection was viewed as a useful tool for acquiring data about the nature of such cognitions, though the methodology fell into disfavor and was largely abandoned during the past century. However, introspective self-reports are still employed in social psychology to assess such constructs as attitudes, leading to continuing debate over the proper role of introspection in scientific psychology.
History of Introspection
The controversial nature of introspection stems from its use as a methodological tool by the structuralists, who sought to create modern, empirical psychology toward the end of the 19th century. Wilhelm Wundt and others trained research subjects to examine and describe their own thoughts in an attempt to create a table of mental elements analogous to chemistry’s periodic table of elements. This method of trained introspectionism was described by Edward Titchener as requiring impartiality, attention, comfort, and freshness. After 40 years of research, structuralists cataloged 50,000 constructs, representing three major classes of elements—sensations, images, and affection—each of which was viewed as possessing four attributes— quality, intensity, duration, and clearness.
The method of trained introspectionism ultimately became bogged down with reliability and validity issues, especially because training inherently colored the reports of introspecting subjects. The approach was criticized by Gestalt theorists, who argued that the overall organization of thoughts is more important than individual elements, and by behaviorists, who argued that behavior, not thought, is the proper focus of scientific psychology. Over the next 50 years, these two approaches dominated Europe and the United States, respectively, and the method of trained introspection was abandoned.
Validity of Introspective Self-Reports
The behaviorist critique calls into question any research method that relies on people’s introspective self-reports of their perceptions, thoughts, or feelings. Yet, such self-report measures are commonly used in social psychology, especially to assess moods, emotions, beliefs, and attitudes, often to good effect. True, concerns are raised periodically that people may distort their self-reports, especially if the attitudes they hold are socially undesirable. And recently researchers have demonstrated that people sometimes hold implicit attitudes of which they are not even aware and which, therefore, cannot be assessed with common self-report measures. One view is that such attitudes reflect an elaborate adaptive subconscious that inherently colors all perceptions, communications, and actions. An alternative view is that implicit attitudes may be relatively rare and frequently overridden by conscious ones.
Critics also argue that introspection necessarily changes the cognitions that people contemplate and report. One program of research suggests that simply thinking about one’s attitudes causes them to become more extreme. Another indicates that thinking about the reasons for one’s attitudes can fundamentally change those attitudes in important ways. For example, in one study, subjects introspected about why they preferred one of two posters before deciding which to take home; others made their choice without introspecting. When contacted weeks later, those who had introspected before choosing were generally less happy with their selection than those who had not. The researchers suggest that introspecting focused subjects on easy-to-communicate justifications for their choice that did not reflect their actual feelings, leading to choices they ultimately found unsatisfying.
One common view is that people are ordinarily better at discerning their own attitudes than they are at introspecting the reasons for, or processes underlying, those attitudes. In one study, shoppers felt several nightgowns, reported which they preferred, and then described the reasons for their preference. In actuality, all the gowns were the same, though people tended to prefer the one on the right, due to a common serial position effect. However, no one correctly reported that their preference was determined by serial position; instead, people made up justifications for their preferences. People’s tendency to introduce theories about their thoughts and preferences, rather than to report such thoughts objectively, underlies many criticisms of introspective methods.
Nonetheless, some psychologists argue that introspection ought to be treated like any other scientific methodology, including modern brain-imaging tasks that may seem more scientific. In other words, researchers need to develop sophisticated theories of the cognitive processes involved in introspection, the factors that affect such processes, and thus the circumstances under which introspection can or cannot provide useful data. In general, introspection is expected to yield more valuable data about the way that stimuli and events are experienced than about the mechanisms or causes of those experiences. And, in general, converging results from several different methods will be more definitive than the results of any one method alone.
Consider, for example, introspective reports of pain. Doctors generally assume that self-reports of the nature, severity, and location of pain are highly informative, even if not totally accurate. When a patient says, “It hurts when I raise my right arm,” this is a key piece of evidence in framing the problem to be addressed and in diagnosing the ailment. Other kinds of data, such as x-rays or brain imaging may also provide useful data, especially when combined with those self-reports. But doctors are much more skeptical of a patient’s speculations about the causes of reported pain, such as “It feels like I tore the bursa.” This is where other methodologies may be more useful. Even so, when the patient has the appropriate knowledge (e.g., she is a doctor herself), even introspections about causation may be valuable. Some writers therefore suggest that refinement of introspective methods may ultimately require that subjects receive special training, a controversial proposal given past criticisms of the method of trained introspection.
- Jack, A. I., & Roepstorff, A. (Eds.). (2003). Trusting the subject: The use of introspective evidence in cognitive science (2 vols.). Exeter, UK: Imprint Academic.
- Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on mental processes. Psychological Review, 84, 231-259.