Ralph L. Smith Mental Retardation Research Center
Pavlov imagined theory as a set of wings: "They allow man to soar...to the heavens. But facts are like the atmosphere against which those wings must beat, and without which the soaring bird will surely plummet back to earth". The metaphor is delicious. Theory is substantial, even beautiful. Facts blow about freely, depending upon imperfect instruments and constantly threatened by unexamined assumptions about what those instruments can validly reveal.
Yet facts remain crucial to the research enterprise, and they have been long- awaited by American audiences of the Vygotskian psychoeducational theory of "zone of proximal development" (ZPD). In the field of mental retardation (MR), ZPD was introduced in clinical and research journals in the 1960s (Luria, 1961; Budoff & Friedman, 1964). But we were never treated to full-length, detailed reports of the empirical work that must have underlain Russia's clinical and research developments of ZPD since Vygotsky's time.
Lacking access to the background studies, we adapted our familiar instruments and simply imagined what a ZPD experiment might look like when using clinical groupings of children as we define them. This led to empirical tests of the ZPD as it relates to static psychometric intelligence. While some Americans studied this relationship using perceptual/cognitive tasks (Campione et al., 1985; Day & Hall, 1988), we used a memory problem that requires deliberate mnemonic activity. We asked, do mentally retarded (MR) children's performance deficiencies in learning, retention or generalization happen because they do not use an appropriate mnemonic activity (strategy), or use it, but poorly. In one study we tested the trainability of MR children's generalization of mnemonic activity (Belmont, Butterfield & Borkowski, 1978). In another, we considered performance within the ZPD.
THE MEMORY PROBLEM. - The memory problem required the child to memorize lists of words for immediate spoken recall. Every list had different words. A computer-controlled tape recorder presented a single word each time the child pressed a button. The child could thereby spend as much time as he wished following each word in the list. As an objective index of memory strategy, the computer recorded the amount of time the child paused after each word (study time). The task was "Circular Recall", meaning that the child attempted to recall the list's last few words followed by its first few words. For example, for a 7-word list, recall began with the last 3 words, then the first 4.
The Ideal Memory Strategy. - The ideal strategy for Circular Recall involves cumulative rehearsal of the list's early words, which will have to be retained until the last few words have been heard and recalled. The last few words are not rehearsed, for they are recalled immediately. Thus, the child can pace himself rapidly through the last few words after thoroughly rehearsing the early words. The computer detects this ideal strategy by the highly characteristic pattern of study times, which increase from word to word during cumulative rehearsal, then become low and flat during the rapid ending.
STUDY DESIGN. - We intended to teach the children to use the ideal strategy, and our main concern was that all children should be working not too far below perfect recall accuracy, so that they would be well motivated.
Pretest. - To achieve this, we gave an adaptive pretest that began with short lists. If the child recalled two consecutive lists perfectly, we increased the list length, continuing until the child reached a level of difficulty at which he failed to give consistent perfect recall.
Training. - The training was done at this failure level. the point was to achieve a modest degree of proximal development, and one happy side-benefit was that the child had a great deal of success while learning to execute the ideal memory strategy. Training involved a two-step Vygotsky sequence. On List 1 we called attention to the easy part of the list (the last few words) and the hard part (the first few words) and suggested that the child spend more time thinking about the hard part than about the easy. No particular kind of thinking was suggested, however.
Starting with List 2, the ideal strategy was explained and demonstrated. For example, for 7-word lists, the child was taught to gather in the first 2 words and rehearse; then add the 3rd word and rehearse all 3; then the 4th and rehearse all 4. The examiner taught the child to pay attention to the rhythmic structure of the rehearsal and to periodically change that structure. In this way, continuous attention was maintained as the rehearsal set grew larger. When the child was confident of recalling the first 4 words, he dashed through the last 3, recalled them immediately, and then recalled the first 4: Circular Recall.
Over the course of eight training lists, every child learned to execute the strategy smoothly, and every child recalled perfectly at least one training list.
A Filler Problem. - Immediately following training on the Circular Recall problem, the children were given five lists of a different memory problem (Serial Recall) without further training. This filler problem could not be solved using the ideal memory strategy taught for Circular Recall.
Retention. - Following the filler problem, unassisted strategy retention was tested using five trials of the original training version of Circular Recall. Alert children would reinstate the ideal strategy.
Generalization. - Finally, for a test of generalization, the children received three trials of modified Circular Recall. For example, if they had been working on 7-word lists and recalling the last 3 words followed by the first 4, for the generalization test they were asked to recall the last 2 words followed by the first 5. This new requirement still responded to the ideal strategy, but the peak rehearsal pause had to occur after the 5th word instead of the 4th. The study-time pattern would clearly reveal such a shift in strategy, indicating adaptive generalization.
RESULTS. - For each unassisted phase in the experiment, we examined the children's study-time patterns to identify ideal strategy use. As a first step, we compared the mean recall accuracy of children who had spontaneously used the ideal strategy (Users) and those who had not (Non-Users).
RECALL ACCURACY
Non-Users Users
MR 2.29 3.68
Gifted 3.00 4.50
For MR and gifted children alike, spontaneous users of the ideal strategy recalled more words than non-users. Given this validation of the ideal strategy's value for these children, the next step was to compute each group's incidence of spontaneous ideal-strategy use at each stage in the study.
INCIDENCE OF SPONTANEOUS USE OF THE IDEAL STRATEGY
Pretest First Hint Retention Generalization
MR .22 .22 .78 .33
Gifted .33 .56 .89 .72
p< .05 p< .05
Few children used the ideal strategy prior to instruction (Pretest). This shows that in spite of the huge difference in their static psychometric intelligence, both groups had much to learn. Static mental age was evidently controlling spontaneous strategy use. Following the first step in the Vygotsky sequence (First Hint), over half of the gifted children used the ideal strategy, while the MR children's strategy use remained at pretest level. The gifted children thus showed significantly more learning potential under partial instruction.
On the unassisted Retention test -- which was given following the filler task -- the vast majority of both groups elected to use the trained strategy. Thus, the MR children showed no evidence of poor strategy retention.
Finally, on the new Circular Recall (Generalization) problem, over 70 percent of the gifted children adapted the ideal strategy to conform to the new recall requirement, whereas only 33 percent of the MR children did so. The gifted showed substantial generalization even as the MR children's strategy selection was virtually collapsing. We conclude that the MR children had no deficiency of strategy retention, but did have significant deficiencies of learning potential and strategy generalization. The latter two findings show that relative to the gifted children, the MR children's ZPD lacked both depth and breadth.
SUBJECTS. - The subjects were 18 MR children 12 to 15 years old with mean static mental age 11 yr. As in Study 1, the basic memory problem was self-paced Circular Recall of 7-item lists, but instead of using words, we reduced the cognitive load to a minimum by using lists of isolated letters of the alphabet. Every list contained 7 different non-rhyming letters. To begin with, the children had to recall the last 3 letters followed by the first 4.
Session 1 began with an unassisted pretest, followed by full training on the ideal strategy (no Vygotsky sequence). This training was followed immediately by an unassisted generalization test like that used in the ZPD analysis: 7-letter Circular Recall, but the number of letters to be recalled at the end of the list changed from 3 to 4, with a corresponding decrease in the early portion from 4 to 3.
All children received 5 lists on this unassisted generalization problem. They were then randomly assigned to 2 groups (N=9/group). One group ("Once Trained") continued the new task without additional instruction. For the other group ("Twice Trained"), we gave direct and detailed instruction concerning generalization of the modified ideal strategy, comparing and contrasting it with the originally trained strategy. This direct instruction regarding the process of generalization was the key instructional component in this study.
One week after Session 1, the children returned for 3 unassisted post-tests given by an examiner who did not know which treatment the children had received in Session 1. The post-tests included a Position Recall problem, in which the child paced himself through the list and was then shown one of the letters. He simply had to point to where in the list that letter had appeared. This problem is formally different from Circular Recall, but it can be solved economically by using the ideal strategy for Circular Recall. In that light, we regard the Position Recall problem as requiring "far" generalization. The remaining two post-test problems were again Circular Recall. They included a variation never before seen ("near" generalization) and a retention test using the original training problem used with both groups.
RESULTS. - As shown in the figure below, the twice-trained children's recall accuracy exceeded the once-trained children's on all three tests, and showed particular strength on far generalization.
Concerning the strategies the children used to achieve these results, the figure shows the 2 groups' average study-time patterns. On the far-generalization problem, only the twice-trained children rehearsed significantly, which explains their recall advantage. Then, on the near-generalization problem, the twice-trained children showed a highly active, perfectly appropriate ideal strategy, while the once-trained group again did no systematic pausing. Finally, on the retention test, both groups showed excellent use of their previously learned strategy, but the twice-trained children put more effort into their rehearsal, and thereby reaped superior recall.
SUMMARY. - In Study 1 and Study 2 the children remembered the trained strategy without assistance, and it did not matter whether the retention test was given following a brief filler problem or following a one-week interval. Thus, retarded children enjoy excellent retention of well-learned, appropriate memory strategies. Since they are clearly slow to learn those strategies, the original training must be well-designed and well-executed.
As a result of the brief generalization training given to half of the children in Study 2, these children made excellent, highly adaptive modifications of the trained strategy on the unassisted generalization tests one week later. We believe that this generalization occurred because our extra training was done using a second, different problem, which permitted us to encourage the children to compare and contrast the two closely-related strategies. That exercise in the comparison of different strategies for doing different tasks was the critical part of the generalization training. We recommend it as a general method for overcoming the retarded child's acute deficiency of generalization.
Belmont, J.M., Butterfield, E.C., & Borkowski, J.G. (1978). Training retarded people to generalize memorization methods across memory tasks. In. M.M. Gruneberg, P.E. Morris, & R.N. Sykes (Eds.), Practical aspects of memory (pp. 418-425). London: Academic Press.
Budoff, M. & Friedman, M. (1964). `Learning potential' as an assessment approach to the adolescent mentally retarded. Journal of Consulting Psychology, 28, 434-439.
Campione, J.C., Brown, A.L., Ferrara, R.A., Jones, R.S., & Steinberg, E. (1985). Breakdowns in flexible use of information: Intelligence-related differences in transfer following equivalent learning performance. Intelligence, 9, 297- 315.
Day, J.D. & Hall, L.K. (1988). Intelligence-related differences in learning and transfer and enhancement of transfer among mentally retarded persons. American Journal of Mental Deficiency, 93, 125-137.
Luria, A.R. (1961). An objective approach to the study of the abnormal child. American Journal of Orthopsychiatry, 31(1), 1-14.
* Dept of Pediatrics, University of Kansas, Kansas City, KS, USA 66160-7337 Phone: 913-588-5982 / FAX: 913-588-5677 / BELMONT@ukanvm.cc.ukans.edu