Consequences of rigid and flexible learning

Linda Baker, John L. Santa, John M. Gentry
1977 Bulletin of the Psychonomic Society  
Three experiments examined the consequences of learning material in rigid and flexible contexts. In Experiment 1, subjects studied and recalled a list of words in serial order. The subjects in one condition received the words in the same order for three trials, while subjects in the second condition received the words in a different random order on each trial. On the final trial, all subjects were required to study and recall the words in a new order. Although the rigid group showed better
more » ... l recall on the third trial, the flexible group was better on the final trial. Experiment 2 replicated these resuhs with a new set of words. Experiment 3 used the same procedure but introduced a completely new set of words on the fourth trial. Similar results were again obtained. Taken together, these experiments suggest that studying material in a rigid context can restrict the flexibility of both learning and retrieval strategies. Several experiments have demonstrated that condi· tions of encoding influence retrieval (Tulving & Osler, 1968; Tulving & Thomson, 1973). Subjects are often restricted at retrieval such that they can only access material via cues encoded at study. A recent series of experiments by Baker, Santa, and Lamwers (Note I) has shown that the restrictiveness of the encoding context depends on the quality of the initial representation. When to-be-remembered words were embedded in wellintegrated encoding contexts, such as sentences or images, external retrieval cues were less effective than if the words had been studied in less integrated contexts. These results lead to the nonintuitive conclusion that good organization may only be beneficial to learning if people are allowed to recall in accordance with their own well-organized strategy. The present series of experiments extends this notion to a serial recall paradigm which more closely approximates conditions of learning in natural situations. During study, students typically go over their notes time after time in the same sequence. Their serial recall of the material would eventually be close to perfect; but what happens if they must organize the information in a new way? The present experiments pursued this question by having subjects study a list of words in either a fixed or random order over a series of trials. On the fmal trial, subjects were asked to learn the material in a new order. We expected that subjects studying material in the fixed, well-integrated context would show better serial recall than the subjects studying the words in variable context. However, on the fmal trial, we expected the fixed-order subjects to exhibit a marked drop because they were forced to break away from their well-learned structure and learn the material in a new way. If the initial learning experience consists of variably ordered lists, there should be no decrement on Carlton T. J ames sponsors this paper and takes full editorial responsibility for it. 58 the final trial. The varied learning experience would have given subjects practice handling the material in a variety of ways. This flexibility should result in better final recall for the flexible condition than for the rigid condition. EXPERIMENT 1 Method Materials. The stimuli consisted of 15 concrete nouns, selected from the Paivio, Yuille, and Madigan (1969) norms. The words were arbitrarily arranged in a list and then three randomized versions were constructed for use on different trials. The words were assigned with the restrictions that no item appear in the same serial position on more than one list and that no two items appear in adjacent positions from list to list. Design. The experiment consisted of a 2 (encoding condition) by 4 (trials) mixed factorial design. Subjects in the rigid condition studied and recalled the words in the same serial order over three trials. Subjects in the flexible encoding condition studied and recalled the words in a different order on each trial. On the fourth trial, all subjects studied and recalled the words in a new serial order. Subjects were 29 introductory psychology students who received course credit for their participation. There were 18 subjects in the rigid condition and 11 subjects in the flexible condition. Procedure. The lists were auditorially presented on tape at a 3-l>ec rate. Four trials were given, in which subjects listened to the list and were then allowed 2 min to recall the words in serial order. Subjects in the rigid condition heard the words in the same order for three trials, and in a new order on the fourth, while subjects in the flexible condition received the words in different orders on all four trials. Results and Discussion The data were scored in two ways, by absolute position and by adjacency. Under the first system, a response was scored as correct if it appeared in the exact position of the input list, independent of other words. The adjacency system scored as correct each pair of words appearing together in output which appeared adjacently in input, regardless of absolute position. Analyses of variance were performed using both measures. Since the pattern of results are quite comparable, only
doi:10.3758/bf03336929 fatcat:kfspkj3plncvra2xh7zjtv4d5m