AN INVESTIGATION OF SELECTIVE REMEMBERING IN AUDITORY SHORT-TERM MEMORY
SECHLER, ELIZABETH S.
Doctor of Philosophy
The research constitutes an investigation of selecting material for remembering, and specifically of the ability to pick out wanted from unwanted items intermixed in the same spoken message. In Experiments 1-4, subjects tried to remember a list of words heard under (a) a no-distraction condition, in which no other words were presented; (b) a precuing condition, in which the to-be-remembered words were intermixed with unwanted words and the to-be-remembered words were specified before list presentation; and (c) a postcuing condition, which differed from the precuing condition in that the to-be-remembered words were not specified until after list presentation. Wanted and unwanted words were distinguished on the basis of semantic category (e.g., colors versus trees). In Experiments 1-3, there were 12 wanted and 12 unwanted words, and the wanted words could be recalled in any order. Recall in the precuing condition was considerably higher than in the postcuing condition and only slightly lower than in the no-distraction condition, indicating a high degree of selective remembering. Interestingly, the degree of selective remembering depended little if any on within-list positions (Experiment 1), rate of word presentation (Experiment 2), or predictability of wanted and unwanted items (Experiment 3). Such results contrast sharply with the suffix effect, in which recall of a list of items is substantially depressed by a single, nominally irrelevant item at the end of a list. Experiment 4 was similar to the first three but incorporated features of the suffix paradigm, namely a short (8-item) list and the requirement that items be recalled in their presentation order. The efficiency of selection, although reduced, was still appreciable. Experiments 5 and 6 sought to reduce the suffix effect by interpolating the suffix item after each to-be-remembered item rather than just the last one, and by presenting the list and suffix items in different voices. The suffix effect was reduced, but only modestly. It is concluded that the last item to be presented is not easily ignored, and that the suffix effect, therefore, represents a special difficulty in selective remembering.