Prepared by Jason Gorscak and James Witkoskie (additions and comments by Frawley)
Outline of reading: "Constraints on the Problem Solving of Persons with Mental Retardation" by Ralph P. Ferretti and Al R. Cavalier (Department of Educational Studies, University of Delaware)
Introduction
The study of cognition in mentally retarded persons has led to three considerations:
1. They are not strategic.
2. Strategy instruction improves performance.
3. Strategies are not easily transferable.
These considerations point to the fact that retarded persons have capacity limitations in working memory.
Problem Solving
Before analyzing exactly how the constraint on working memory affects performance one must look at what exactly problem solving is. A problem is a difference between the current state and the goal state. Problem solving is moving from the current state to the goal state. Problem solving is accomplished by activating several sub-processes that solve part of the problem and move the person closer to the goal state. These sub-processes are controlled by a more general process, the "self management processor."
The essence of problem solving is the processing of data. There are three procedure for information processing that are used:
1. Activate long term memory to access information that one already has
2. Organize information into a form that is easier to store
3. Automate the sequence.
This can be done if one uses controls to cut down on the number of variables.
Note: assuming the mind is a computing device, there can this be both software and hardware issues.
Software issues: representation, routines and subroutines, executive processes
Hardware issues: memory, efficiency, and control
Note how these two sets of issues lead to differences in causes of breakdown in problem solving. Are there representational deficits, impairments in the content of data structures? Are there impairments in resources and capacity for holding the elements of a problem space in memory? Are there executive software deficits? Are there deficits in automating processes?
Retarded persons are believed to have a lack of ability to use self-management processing in problem solving because they do not have the memory resources necessary to store the self-management programs and the information used on the specific task at hand. They are unable to transfer procedures from one task to another. This theory is supported by the fact that retarded people's abilities to problem solve improves when they are taught methods of self-management. However, there is no conclusive evidence because it has not been shown that self-management is even used effectively by individuals of normal intelligence in problem solving.
The self-management program involves six steps:
1 Define Goal
2 Select Strategy
3 Implement Strategy
4 Monitor Progress
5 Evaluate Strategy Based on Progress
6 Revise Strategy If No Progress
Cognitive Training Studies
To show that self-management is necessary to perform problem solving and that retarded persons lack this ability of self-management, several experiment were performed. These experiments involved teaching retarded people to perform self-management tasks.
One experiment involved teaching recall. The participants in the experiment were taught to anticipate the name of an object after one run through the list or keep repeating the name of the object after they were introduced to it. The test subjects were able to learn the objects quicker than the control group. However, when these subjects were then tested on identifying pictures, which uses the same method as the object identification, the retarded persons and young children had to be prompted to use these methods that they were taught. Retarded persons and young children also did not retain the method as long as more mentally mature persons. This shows their inability to transfer.
These examples do not totally support the concept of self-management, however, because researchers have failed to separate all the variables from the experiments. But the conclusion can be drawn that after external controls are used self-management is successful. Target behavior in the subjects was created before the self-management was taught and external consequences for target behavior were employed before the techniques of self-management were taught. Several other methods of self-management were tested with relatively consistent results.
There are two possible explanations for the change in behavior (reactivity) that is caused by self-monitoring training, but it is tough to know if the test subjects are able to use objective judgment in these self evaluations that produce the change. (1) A feedback loop might form. (2) Self-evaluation and self-consequence might increase due to salient cues.
This does not completely prove the theory of self-management but it does give it support. If self-management really is a necessary feature in problem solving and it is limited by the amount of memory available then it would put constraints on the type of problems that could get solved, such as the complexity of the tasks. It would also limit one's ability to use self-management since the self-management program also uses memory.
Working Memory and Strategy Production
Other empirical evidence that supports working memory's affect on problem solving comes from the following:
1. Retarded Persons perform close to normal subjects when their memory resources are not overloaded. (Material is not too difficult)
2. They perform close to the norm when comprehension is good. (No superfluous directions). It is important to note that the retarded persons are not as likely as normal persons to ask questions about things that they do not understand. They need to be prompted about everything.
3. They perform well when they have the content knowledge.
Retarded children fail to use the skills outside of their initial situation that they learned the skill in. This is called welding. However, children who were instructed with general rules performed better on the transfer that children who just learned the specific cases through trial and error.
Cross example applications show how general memory can accelerate task specific learning. Children were able to apply general principles to specific tasks rather than the other way around Also, the closer related two tasks were the more easily knowledge from one task could be applied to the other. The example used was the application of marbles rolling down an inclined plane and a balance with weights on it. There were three groups of children. One group was just taught the scale. Another group was taught a general unrelated problem and then the scale. Another group was taught the plane and then the scale. The group that learned the plane problem first excelled at learning the other problem.
Conclusions
The experiments are inconclusive because they did not control all the stimuli. However, retarded persons are productive in certain situations and can sometimes perform procedure transfer although not as well as normal subjects. There is the possibility that there is a link between working memory capacity and learning.
There are restrictions to retarded persons' capabilities. This was shown by Campione's matrices experiments. Other support for this theory is the weights problem that was also used in the transfer tests. Retarded persons performed as well as normal persons when only one rule was used such as distance or weight. However, when asked to combine these two rules, the retarded persons fell well below the norm. Retarded persons also perform better when given multiple examples. It helps them transfer procedures. However, they often need prompting to transfer. Retarded persons do not have the working memory capacity to access these learned procedures while keeping track of the data of the problem at hand and transferring the procedures.
Working memory is determined by several factors. Functional memory limits representations of information and therefore hinders the acquisition of solutions to problem. A theory of how working memory is restricted is the processing time might be slower which causes a back-up. ------------------------------------------------------------------------ Short additional summary on class discussions on mental retardation on November 20-25, 1996.
Mental Retardation is sub-average intellectual functioning, as defined by IQ, existing concurrently with limitations in two or more adaptive skill areas. There are both familial and organic causes of MR. Familial retardation is derived from outside sources, family, society, etc. Organic retardation is a functional code problem inside the mind-brain; organic bases tend to result in much lower overall IQ.
Mentally retarded persons' problem solving relates to passive memory, strategic passivity, failure to maintain and transfer strategies, and metacognitive deficiencies. It seems that the mentally retarded have the strategy to solve problems, but they do not "know" how to retrieve, access, or control the knowledge from their mind-brain's "storage bin." Mental retardation appears to be related to problems in working memory, for example in the capacity & resources. But some mentally retarded lack domain specific knowledge too.
Further problems with the idea that MR is a problem of access, control, and domain: (1) often it is difficult to sort out metacognitive information from domain knowledge; (2) metacognitive knowledge is poorly correlated with strategy use; (3) tests of metacognitive knowledge are problematic; (4) strategic deficiencies can be found in all sorts of populations, not just MR.
So how do we study effective problem solving? There are at least five factors to consider:
1. elaborated and connected domain knowledge
2. relevant pattern detection that lead to effective outcomes
3. strategies anchored to authentic tasks
4. strategies are effective under certain conditions
5. working memory has limits
Focus on the last since that has clear tests and effects:
1. MR individuals are slow (in scanning, e.g.)
2. But young non-MR are slow, too! 3. Speed differences in performance are associated with span in working memory
4. capacity differences are related to strategy use
5. strategies involve decomposition and so strain working memory
In divided attention tasks -- which affect working memory by making the aggregate demands exceed resources -- MR individuals show impairment. Hence, working memory deficiency seems to be characteristic of MR problem solving. This leads to specific recommendations for intervention: give instructional support that help working memory; learning and strategy transfer should take into account the demands made on working memory since MR individuals appear to benefit from help.
Question: if MR individuals so benefit, is working memory impairment a performance problem? Compare the behavior of Williams syndrome individuals, who cannot improve their spatial performance -- hence a competence impairment?