Optional shift behavior in children and young and elderly adults

Optional shift behavior in children and young and elderly adults KENNETH L. WITTE University of Arkansas, Fayetteville, Ark. 72701 Optional intradimensional and extradimensional shift behavior in preschool children, college·aged adults, and two groups of elderly adults was investigated in this study. The young adults made reliably more intradimensional shifts than did the children. This finding is in accord with the results of earlier studies in which optional reversal and nonreversal shift behavior has been investigated as a function of age. While the two elderly groups did not differ reliably in their shift behavior, the performance of the younger elderly Ss was comparable to that obtained for the college Ss, while the performance of the older elderly Ss was c9mparable to that obtained for the children. The optional shift paradigm has been used by the Kendlers (Kendler, Kendler, & Learnard, 1962; Kendler & Kendler, 1966, 1970) for examining the development of mediational processes in children's discrirnination learning. This research has consistently indicated that between preschool and college age, the percentage of Ss making an optional reversal shift increases with age, while the percentage of Ss making an optional non re versal or nonselective shift decreases. On the basis of these findings as well as other data (e.g., Kendler & Kendler, 1959), the Kendlers have proposed a developmental mediation theory which assumes an ontogeny of children 's shift behavior. The Kendlers' mediation theory can be contrasted to alternative models that hold that mediation is not related to the phylogenetic or ontogenetic level of the S. For example, Zeaman & House (1963) hold that discrimination learning requires a chain of two responses. The first is a central mediating response to the relevant stimulus dimension; the second is an instrumental response of approach to a particular dimensional cue. This view holds that discrimination leaming is always governed by mediation and that reported ontogenetic differences are due, in part, to artifacts involving procedural factors (see, e.g., Dickerson, Wagner, & Campione, 1970). Campione (1970) recently obtained evidence which he interpreted as being inconsistent with the Kendlers' notion that there exists an ontogeny of shift behavior. He employed the total'change optional'shift paradigm to exarnine the behavior of preschool and second-grade children and found that the children's tendency to make optional intradimensional shifts was independent of age level. However, th is evidence is not necessarily inconsistent with the Kendlers' Psychon. Sei., 1971, Vol. 25 (6) developmental theory in that they have suggested that the probability that children's behavior will be mediated gradually increases with increasing age. Thus, Campione's failure to find age differences may have resulted because the age difference between his two age groups was too small. To investigate this possibility, a greater age difference-preschool vs college age--was used in the present study. It was hypothesized that the college-aged Ss would make reliably more öptional intra dimensional shifts than would the preschool children. A second purpose of the study was to examine the shift behavior of elderly adults. Although age differences between young and old adults in the use of mediators in paired-associate learning has been reported (Hulicka & Grossman, 1967; Hulicka, Sterns, & Grossman, 1967), little attention has been focused upon possible age differences in the various shift paradigms (Botwinick, 1970). Crovitz (1964) found that only 10 of 26 aged men (mean age, 65.1 years) were able to learn areversal shift, whereas 28 of 32 young Ss (mean age, 23.8 years) learned the problem. She suggested that aged Ss may be comparable to young children in terms of a "mediational deficiency." In a study exammmg the reversal, intradimensional, and extradimensional shift behavior of elderly men (mean age, 66 years), Nehrke & Coppinger (1971) obtained evidence which led them to suggest that their Ss were functioning in a manner simiIar to children in transition from mediational deficiency or single-stage theory to a mediational level. On the basis of the preceding results, it was predicted that the optional shift behavior of the elderly Ss would parallel that obtained for the children. SUBJECTS The Ss included 32 preschool children, 32 college students, and 32 elderly adults. The data for 4 additional college Ss were discarded because of experimental error; no S was excJuded for failure to learn either the initial or the optional shift discrirnination. The children were obtained from two private preschools and ranged in age from 36 to 61 months, with a mean age of 52.4 months. The college Ss were obtained from general psychology classes at the University of Arkansas and ranged in age from 18 years, 7 months to 25 years, 3 months, with a mean age of 19 years, 9.2 months. The older adults were obtained from an adult center in Fayetteville, Arkansas. Their ages ranged from 53 years, 0 months to 81 years, 0 months, with a me an age of 69 years, 0.2 months. All but three of these Ss had completed high school, and 14 had so me college training; they all reported that they currently were in good health. STIMULI Colored geometric forms pasted on 5 x 8 in. white cards served as stimuli. There were a total of 64 color-form patterns resulting from the factorial combination of eight forms (square, circJe, cross, triangle, diamond, star, X, and T) and eight colors (red, blue, gray, green, black, brown, orange, and yellow). Maximum height of the forms was 4.75 in., while the maximum width was 3.75 in. Sixteen different sets of problems were generated from these stimuli, and two Ss from each age level were tested with each set. PROCEDURE Each S was tested individually. The stimulus cards were kept behind a small screen; E initiated a trial by manually presenting a given stimulus card. The S pointed at the stimulus he thought to be correct; a noncorrection procedure was used throughout the experiment, and the Ss' responses were verbally reinforced by E. The intertrial interval was approximately 5 sec. On the initial discrimination, half the Ss within each age level were trained with color relevant, while the remaining Ss were trained with form relevant. The irrelevant dimension varied within trials. The spatial arrangement of the stimuli on each trial was deterrnined by a Fellows' (1967) series. On both t he i ni tial and 0 ptional shift discrirninations, each cue was positive and negative an equal number of times within each age group. Training was continued on the initial discrimination to a criterion of 10 successive correct responses. After the S reached criterion on the initial problem, the optional shift discrimination was introduced with no ostensible break in the procedure. For each S, entirely new color and form stimuli were presented, and Ss were 329 required to learn a discrimination (...truncated)