Effects of Systematic Variations in Reinforcement Contingencies on Learner Performance

A previous experiment (Sullivan, Baker, and Schutz, 1967) suggests that the reinforcing properties of the feedback provided by immediate knowledge of results may strengthen behavior other than that sought by the instructor. Subjects (Ss) receiving immediate feedback for responses to en route mastery-test items inserted at various points within textual material scored significantly lower on the mastery items than did Ss receiving no knowledge of results. "Feedback" Ss spent considerably less time than "no feedback" Ss studying the textual material. They neglected the textual material to some degree and attempted instead to utilize the instructional value of the immediate feedback following their responses to the mastery items. That the "feedback" Ss were successful in learning from the immediate knowledge of results is suggested by the fact that their subsequent performance on the end-of-program criterion test was comparable to that of the "no feedback" Ss, even though their performance on the en route mastery tests was significantly inferior. These data suggest that if one is to capitalize maximally upon the advantages of instructional refinements such as the immediate feedback procedure, greater control of the textual material over learner responses must be established and maintained.