An analysis of recall and recognition processes in verbal learning

The purpose of this experiment was to demonstrate that the superiority of recognition over recall is due in part to the use of the whole item as the basic unit of measurement. About some items, S s learn less than the whole item, and it is on the basis of this partial learning that items which cannot be recalled are, nevertheless, easily recognized. Lists of items of three different orders of approximation to English were given to S s to learn. Retention was measured after each learning trial either by recall, standard recognition, or restricted recognition. The standard recognition test contained the original items plus an equal number of other items from the same order of approximation. In an effort to restrict the potential effectiveness of partial learning, the other recognition test contained incorrect alternatives differing from the original items in only one letter. Results of the experiment showed that when opportunities for partial learning to be of use in recognition were restricted, much of the difference between recall and recognition disappeared. Order of approximation and its interaction with method of measurement also had statistically significant effects.