When can we introspect accurately about mental processes?

Abstract

Some theorists have suggested that the cognitive processes determining a person’s perforance in a given task are unconscious, making introspection a poor research tool for studying these processes. Others have argued that the relevant processes usually are consciously conrolled and can be detailed by asking the person to introspect. Here, a synthesis of these two positions, a dual-factor approach, is proposed. Some of the processes involved in achieving a cognitive goal, such as learning a new concept, are viewed as unconscious and automatic; howver, other processes are intentionally allocated conscious attention, in certain tasks, to accomlish othe goal. To illustrate this dual-factor position, evidence is presented in support of the view that when concept learning occurs solely by automatic frequency processing, introspective reports are inaccurate, but when the nature of the task prompts intentional hypothesis testing, introspective reports are accurate, revealing clues that subjects engage in a conscious hypothesisesting strategy.