Forgetting of word associates as a function of recall interval

Abstract

A series of experiments is reported which sought to replicate the observations of Parkin et al. (1982) that recall of self‐generated associates of emotional words is poorer than recall of associates of neutral words on immediate testing, but that the reverse is true when testing is delayed. In the first experiment independent groups of subjects were tested for recall immediately after administration of the word association test used by Parkin et al. and after an interval of one day. Associates of emotional words were less well recalled than associates of neutral words at both time intervals. In the second experiment different stimulus materials were used in an attempt to control for a number of possible confounds in the word list used in the first experiment. Independent groups were tested for recall immediately and after intervals of one hour, one day and one week. Results indicated poorer recall for associates of emotional than neutral words at all test intervals. A third experiment examined attributes of the words given as associates in Expt 2 and demonstrated that the major difference between the associates of emotional and neutral words was their affective value. A possible interpretation of the influence of emotion on forgetting was offered in terms of network theory of human memory.