Postcriterion overlearning reduces the effectiveness of the method of adjusted learning

One goal of the method of adjusted learning (“dropout procedure”) is to equate the strength of different Items for subsequent retention. Some studies have yielded results in accord with this goal while others have not, with no known reason for the discrepancy. We manipulated the number of postcriterion overlearning trials and found that the adjusted learning procedure was effective in equating items for subsequent retention when learning terminated after one correct response per item but ineffective when postcriterion overlearning trials occurred. This finding is especially important for those versions of the adjusted learning procedure, such as the frequently utilized variant developed by Battig (1965), that incorporate overlearning of the to-be-equated items.