Abstract
The principal purpose of this paper was to test the adequacy of the informational function of verbal announcement “right” “wrong” and “nothing” in paired-associate learning. This informatioal function may consist of both the information value of “right” and “wrong” and the information concerning which item is to be learned and which one has been learned.
Exp. I was designed to test the Buchwalds theory which suggested that the effects of verbal announcement of “right” and “wrong” was fully explained only by the informnational function. He postulates that there are separate memory processes for responses and for outcomes. Further, the outcomes. are assumed to have no influences on the stimulus-response associate process. The procedure of our experiment involved immediate outcome or delayed outcome of “right” or “wrong” using the Thorndike's paired-associate task. The Ss were given two trials with response alternative of 4 or 6. On immediate outcome condition, E said “righ”,“wrong” or “nothing” immediately after S's response. But on delayed outcome condition, E informed S that his response had been “right”,“wrong” or “nothing” when the same item was presented on the second trial. The obtained results showed that delayed outcome of “wrong” led to significantly less response repetition than all other conditions. Differences among other conditions were generally not significant. These results may be interpreted as follows: If the outcome is given immediately after the response, it may be forgotten when the item is presented next time. If the outcome of a previously given response is not presented again, it will insure that the information will be available at the point at which it is needed.
Exp. II was designed to investigate the effects of the information concerning which item was to be learned using the study-test paired-associate task. The design of this experiment was a 2×2×2 factorial one, in which a between factor was a length of list (6-and 10-word lists) and two within factors were the announcement of “right” and “wrong” and the trials required to attain to the two criteria. 00≤p≤.50 and. 50<p≤1.00. Two different lists were presented to the same S under the two different conditions for the announcement of “right” and “wrong”. That is, S was informed that his response had been “right” or “wrong” when the item was presented on the next study trial. Although the information value- of “right” was made equal to that of wrong in this situation, the information, concerning which item was to be learned may have been different by the requirement of memory task. More specifically, the announcement of “right” may be important at the beginning of study trials, the announcement of “wrong” may, however, be important at the end of study trials. The obtained results revealed significant, second-order interaction. This interaction may be interpreted as follows: With regard to the 10-word list, the announcement of “right” had a tendency to be significantly superior to that of “wrong” through the trials. Regarding the 6-word list, the announcement of “right” was significantly superior to that of “wrong” at the beginning of trials, while it was significantly inferior to that of “wrong” at the end of trials. Therfore, the above fact and its interpretation do not support Mosberg's hypothesis which suggests that the relative importance of “right” and “wrong” may change at. 50 correct proportion. Finally, the relation between the informational function and rehearsal buffer model was discussed.
