Guidance Hypothesis

The guidance hypothesis is an account in motor skill learning research holding that augmented feedback about the outcome of a movement, commonly called knowledge of results, functions much like an external guide that helps a learner perform accurately and correct errors while that feedback keeps being provided, but that heavy reliance on this externally supplied information can prevent the learner from developing the internal error-detection and self-correction processes that are needed to perform the skill well once the feedback is no longer available. This account explains a range of findings in which a practice schedule that provides very frequent guiding feedback produces good, accurate-looking performance during the practice sessions themselves, yet paradoxically leaves weaker long-term retention of the skill than a practice schedule that provides feedback less often, since the less-guided learner is forced to rely on, and thereby strengthen, their own internal sense of whether a given attempt was correct. The hypothesis was articulated most influentially by the psychologists Alan Salmoni, Richard Schmidt, and Charles Walter in an extensively cited 1984 review of the knowledge-of-results literature, and it has since become a standard framework in motor-learning textbooks for explaining why several different feedback-reduction techniques, including the bandwidth and summary knowledge-of-results effects documented elsewhere in this section, each tend to improve retention despite objectively providing the learner with less total feedback.