FORGETTING HAS LONG BEEN ATTRIBUTED to retroactive inhibition (RI), with a minor role assigned to proactive inhibition (PI), on the basis of a wealth of evidence from studies of short-term retention. Underwood (1948a) found, however, that the amount of PI and RI became equal after a longer retention interval (48 hours). In addition, he has since suggested that, except for conditions in which RI is experimentally produced, PI may account for the greater proportion of all forgetting (Underwood, 1957). The question then arises as to how the interference of prior learning on recall is affected when interpolated learning is also introduced. Forgetting virtually always occurs in the presence of some prior and interpolated learning.Three recent studies (Hall & Ugelow, 1959;Seidel, 1959; Tulving & Thornton, 1959) have introduced various amounts of both prior learning (PL) and interpolated learning (IL) in the same conditions. It is doubtful that results either of Hall and Ugelow or of Tulving and Thornton are relevant to combined proactive and retroactive effects. Neither study employed procedures known to be conducive to reliable demonstrations of PI, and little (if any) PI resulted in either.Seidel used serial learning (nonsense syllables), PL 24 hours prior to the critical learning (CL), a 24-hour retention interval, and IL immediately prior to recall. He found what appeared to be a simple summation of the two inhibiting effects.The purpose of the present studies was to test for an interaction of PL and IL, using paired associates, a short retention interval (30 minutes), and PL, CL, and IL in one continuous session. Seidel's introduction of IL immediately before recall may have resulted in quite different effects than would appear with the more conventional proximity of IL and CL. EXPERIMENT I Subjects and MaterialsTwenty-four paid Ss, twenty-one male and three female university students, learned lists of ten paired-associate two-syllable adjectives. These were presented on a Gerbrand memory drum with a 3-sec. anticipation period and 9-sec. interval
ColumbiaVOEKS'S (1948) FORMULATION of the postremity principle asserts that the most probable response on trial n is the response made upon die last preceding occurrence of the stimuli at that choice point, generally on trial n-1. In three manual-maze learning tasks (Voeks, 1948) individual response predictions based on this recency principle were superior to those based on frequency of past occurrence and were far more accurate than would be expected on the basis of chance. 3 The lack of perfect postremity effect Voeks attributed to (c) difficulty in recording the actual Tast-made" response, and (b) changes in the stimulus situation at the choice point from trial to trial. When the lastrecorded response is "incorrect," the signal indicating the incorrectness of the response may elicit additional covert (unrecorded) responses in the presence of many of the same stimuli to which the incorrect response was made. Thus predictions based on the last-recorded response will be inaccurate.Voeks considers changes in proprioceptive stimuli, arising from postural changes and varied movements, of primary importance since she, following Guthrie (1935), believes that these stimuli comprise a large part of the stimulus situation which becomes associated with a response. Changes in excitement and curiosity (especially on early trials) and changes in external stimuli further augment the stimulus instability.In a later study on eyelid conditioning, Voeks (1954) attributed the greater success of postremity predictions to the unusually careful control of many of the above sources of stimuli. Waters and Beitz (1950), however, found no effect on the accuracy of postremity predictions when they introduced postural and movement changes from trial to trial, compared to usual training conditions. Apart from the fact that the findings were limited to one stage of learning, both of their conditions would be 1 Tbis report is largely based on a thesis submitted by the senior author to the University of British Columbia in partial fulfilment of the requirements for an MA. degree. Part of the material in this report was presented in a paper at the Canadian Psychological Association meetings, June, 1957.2 Now at the University of Manitoba. s"Chance" in this paper refers to the expected probability of a response on the basis of random guessing, i.e., 1/n, where n is the number of response alternatives. Observed frequencies of responses above or below this level, evidence of response selection, may be interpreted in a number of ways. This problem will be alluded to later in the paper.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.