Role of stimulus-term and serial-position cues in constant-order paired-associate learning.

Abstract

72 Ss first learned 12 CVC pairs presented in the same serial order on every trial. Half of the Ss learned an LH list (low formal stimulus, S, term similarity, high formal response term similarity), and half learned an HL list. Then a 2nd transfer list was learned to evaluate the relative cue strength of S-term and serial position (SP) cues in different positions of the 1st list. For S-terms this was accomplished by changing the order of the Ist-list pairs and for SP cues by changing only the order of the S-terms. The transfer results showed a greater utilization of S-terms with List LH but a greater utilization of SP cues with List HL. Under both similarity conditions, however, better performance was displayed for S-terms taken from the middle of the 1st list, while SP cues facilitated 2nd-list learning more in beginning and end positions. Paired-associate (PA) learning experiments usually vary the order of presentation of the pairs on every trial. Consequently, 5 must rely upon the elements of the stimulus (S) term to cue the correct response (R) term. However, when the pairs of the list are presented in the same serial order on every trial, 5 can also respond to whatever cues are tied to serial position (SP), such as ordinal position, relative position, or preceding R-terms in the list. Thus, under a constantorder method of pair presentation, 51 is provided with a compound stimulus consisting of S-term and SP cues. In a study designed to investigate the relative importance of these two types of cues in learning, Newman and Saltz (1962) had 5s first study a list of pairs in serial order. Then a recall test was given in which the serial order of the S-terms was the same or different from that used during study trials.