THE INFLUENCE OF CONTEXT UPON

LEARNING AND RECALL

BY SHUH PAN

Psychological Laboratory, University of Chicago

INTRODUCTION

This experiment was designed to test the dependence of learning and recall upon the environmental situation in which the material is presented. Our subjects were required to memorize pairs of words presented in various contextual situations, and they were later tested for their ability to recall this material when the contextual situation was present, absent, or altered in certain specified ways.

A somewhat similar experiment has been performed with the use of animals under the maze problem. After his rats had mastered the problem, Watson (6) rotated the maze in reference to its extraneous environment, and then tested the ability of his animals to run the maze with this change of environmental conditions. The experiment was repeated by Porter (4) and by Hunter (2), using sparrows and pigeons respectively. In all three experiments the animals were confused, the confusion being indicated by an increase in the time and in number of errors. Presumably the animals had learned to respond to some extent to the sensory environment in which the maze was situated, and hence they were disturbed when this relation of the maze to its environment was altered.

This dependence of the 'maze habit' upon the extraneous environment has been systematically studied by Carr (1), who varied the relation of the maze to its environment in various ways while the rats were learning and after they had mastered the maze. His experimental conditions and results are listed in Table I. On the basis of these data Carr concluded that the rat learns to adjust to the entire sensory situation and hence stability of this situation is conducive to learning and to the functioning of the act after it is learned. "Any sensori

motor act cannot be regarded as an isolated independent function; the act was learned within a wider sensory environment, and it never ceases to be wholly free from those conditions either during or after its development" (p. 291).

Smith and Guthrie (5) mention two experiments in which human subjects were employed. In one of these experiments the subjects learned a list of 10 nonsense syllables in the laboratory, and relearned it 72 hours later in the laboratory; learned a second list in the laboratory and relearned it out of doors; learned a third list out of doors and relearned it in the laboratory, and, finally, learned a fourth list out of doors and relearned it in the same environmental situation. "In eight of the ten subjects there was greater saving in each case where relearning occurred in the same surroundings in which the first learning had taken place. Two subjects showed in one of their four series a greater saving where relearning had occurred under dissimilar conditions. An average of 11.4 per cent. fewer repetitions were required to relearn in similar surroundings" (p. 112). In their second experiment the subjects copied a list of letters on a typewriter the keys of which had been arranged in a random order. "A variable condition consisting of the presence or absence of the odor of oil of peppermint was used." A second practice occurred 24 hours later. "Thirteen subjects were used, and the average saving in the time required to repeat the list was 9.4 per cent. greater when the repetition took place under the condition of the first practice" (p. 113).

Another study with human subjects is reported by Wong and Brown (), who had two similar groups of subjects solving Yerkes' multiple-choice problems in two rooms; a 'good' room and a 'bad' room. The 'bad' room was an attic room, poorly lighted, dusty, and filled with various objects in disorder. The 'good' room was a well-lighted and well-furnished office. In the given time allotted, the subjects working under the distractive condition solved an average of 5.6 problems per individual, and the average number of trials required per individual was 230. The corresponding records for the other group were 7.8 and 243. The group working under a favorable

environmental condition was thus able to work faster and solve more problems in a given amount of time.

MATERIALS AND METHOD

The materials to be memorized, with one exception, consisted of lists of 20 pairs of logically unrelated words of common usage. Words of one or two syllables were used for the most part. Thirty-two such lists were employed. With a few exceptions, no word was used twice with the same subject.

These lists were presented by means of a memory drum. The time of exposure per pair of words was 3 seconds. Three or four presentations of a list were given in succession, and the amount learned was tested by an immediate written reproduction. The material was thus only partially learned.

The amount retained after 24 hours was tested by a similar written reproduction. In these tests, both immediate and delayed, the stimulus words were presented in an altered temporal order, and the subjects were required to write the response word upon a blank card provided for that purpose. The time for recall was not limited, but it did not exceed 10 seconds a pair.

In a subsidiary experiment, the material consisted of a list of 24 pictures of human faces, each paired with an unfamiliar name. These pictures were taken from a popular magazine, the two sexes being equally represented. Each face and its paired name were presented on a picture post-card. The 24 cards were arranged in a pack, face down. The cards were turned and exposed one by one. The exposure time per card was 3 seconds. The subjects were tested for the amount learned after each presentation of the pack, and the presentations were continued until the material was perfectly memorized.

Retention in this case was tested after 48 hours. To test retention as well as learning, a duplicate pack of cards without the names was employed, and the order in which the cards were presented was altered. The names were orally reproduced. The time for recall was not limited, but it did not exceed 10 seconds a card.

All reproductions for both types of material were scored in terms of the number of verbal associations correctly reproduced. Errors in spelling and pronunciation were ignored in so far as it was evident that the proper word was intended.

Fifty-six subjects served in the experiments, but no subject memorized all of the lists. Thirteen of the subjects were graduate students in the Department of Psychology, and the remainder were Chinese students recruited from various departments.

These materials were presented in a given environmental context. Three kinds of context were employed: words, numbers, and pictures.

The picture on the post-card upon which the face and name were placed constituted the context for this type of material. All of the pictures represented parks, buildings, statues, and other public places in Chicago.

A verbal or numerical context was employed for the pairs of association words. The pair of words to be memorized was always typed in capitals, while the words or numbers constituting their context were always typed in small letters. Usually the context of each pair consisted of two words, one placed above and one below. In a few lists, only one word was employed for the context because of the difficulty of securing suitable words. Only one number was used, and this number corresponded to the position of the pair in the order of presentation. When a number or one word was used, it was written above the pair of associated words. The contextual words bore one of several relations to the pair to be memorized. The kinds of context used are illustrated in Table II. The contextual words may bear no logical relation to either of the members of the pair; they may be logically related to the response word; logically related to the stimulus word; or logically related to both stimulus and response words. Each pair of a list has its own context, and this particular contextual relation always remains constant during the period of learning, unless otherwise specified. The various pairs of associated words composing any list always have the same kind of context.