Repetition and Paired Associates Learning
Alfred B. Kristofferson · Science · 1961
Recent experiments have suggested that paired-associate items which are practiced and missed are equivalent to new items on future trials, supporting an all-or-none interpretation of the formation of associative bonds. Experiments reported here show that if correct items are discarded after each trial the probability of getting a missed item correct on the next trial is higher than the probability of getting a new item correct.