Cognition Research Methods

Introspection

In Chapter 1, we discussed some of the limits on introspection as a research tool, and, in fact, our discussion throughout the textbook has rarely relied on introspective evidence. This is because, as one concern, introspection relies on what people remember about their own mental processes, and we cannot count on these memories being reliable. In addition, introspections are usually reported verbally: The person uses words to describe what happened in his or her own mind. But as we discussed in Chapter 10, some thoughts are nonverbal in content and may not be captured adequately by a verbal description.

As a further problem, your introspections, by definition, involve an “inspection” of your mental life, and so this method necessarily rests on the assumption that your mental state is “visible” to you. (You can’t inspect something that’s invisible!) As Chapter 13 describes, however, a great deal of mental activity goes on outside of awareness, and so it is, in fact, “invisible” to introspection. This provides yet another limit on introspection as a source of scientific evidence—because introspective data will necessarily be incomplete in what they tell us about mental processes.

Let’s be careful, though, not to overstate these claims, because unmistakably, introspective reports sometimes do have value. For example, in the study of problem solving, researchers sometimes ask people simply to “think out loud” as a means of discovering what strategies the people are using as they work on the problem. Likewise, in Chapter 10 we acknowledged the complexities attached to someone’s introspective reports about the vividness of his or her own mental images, but we also argued that these vividness reports can be an important source of data about images. (Also see Demonstration 13.2.) And in Chapter 6, we explored the nature of implicit memories; an important source of data there was people’s introspective reports about whether a stimulus “felt familiar” or not.

How can we reconcile these uses of introspective data with the concerns we’ve raised about introspection? How can we argue that introspective data are of questionable value but then turn around and use introspective data? The answer lies in the simple fact that some thoughts are conscious, memorable, and easily verbalized; for thoughts like these, introspection can provide valuable data. The obvious challenge, therefore, lies in determining which thoughts are in this category—and so available for introspectively based self-report—and which thoughts are not.

How does this determination proceed? Let’s say that “think out loud” data indicate that participants are relying on a certain strategy in solving problems. We then need to find other evidence that might confirm (or disconfirm) this introspection. We can ask, for example, whether people make the sorts of errors that we’d expect if they are, in fact, using the strategy suggested by the self-report. We can also ask whether people have trouble with problems that can’t easily be solved via the strategy suggested by the self-report. In these ways, we can check on the introspections and thus find out if the self-reports provide useful evidence.

Likewise, in Chapter 10 we discussed some of the evidence indicating that self-reports of image vividness do have value. Specifically, we described evidence that reveals a relationship between these reports, on the one hand, and how well people do in certain imagery tasks, on the other hand. Other evidence indicates a link between these imagery self-reports and activation levels in the visual cortex. So here, too, we can document the value of the introspective evidence by checking the introspections against other types of data, including behavioral data and data from neuroscience.

The point, then, is that introspection is neither wholly worthless nor wonderfully reliable. Instead, introspection can provide fabulous clues about what’s going on in someone’s mind—but we then need to find other means of checking on those clues to determine whether they are misleading. But let’s also note that introspection is not unique in this regard. Any research tool must prove its worth—by means of data that in one fashion or another validate the results obtained with that tool. (Consider, for example, our discussion in Chapter 12 of the steps needed to assess the validity of intelligence tests.) In this way, we use our research methods to build our science, but we also use our science to check on and, where possible, refine our research methods.

Critical Questions

1.
fiogf49gjkf0d
What are some potential problems with the use of introspection as a tool in cognitive psychology?
2.
fiogf49gjkf0d
For what sorts of questions is introspection a potentially useful tool? How can we use introspection wisely, despite its limitations?
3.
fiogf49gjkf0d
Choose an experiment discussed in your text or in class in which participants used introspection, verbally described a strategy, or performed an explicit task. Describe a hypothetical follow-up experiment that could complement, strengthen, or potentially disprove the first experiment’s conclusion.

Submit to Gradebook:

First Name:
 
Last Name:
 
Your Email Address:
 
Your Professor's Email Address:
 
Section: