Inflexibility of experts—Reality or myth?

[watertank]

How does the knowledge of experts affect their behaviour in situations that require unusual methods of dealing? One possibility, loosely originating in research on creativity and skill acquisition, is that an increase in expertise can lead to inflexibility of thought due to automation of procedures. Yet another possibility, based on expertise research, is that experts’ knowledge leads to flexibility of thought.
[the authors} tested these two possibilities in a series of experiments using the Einstellung (set) effect paradigm. Chess players tried to solve problems that had both a familiar but non-optimal solution and a better but less familiar one. The more familiar solution induced the Einstellung (set) effect even in experts, preventing them from finding the optimal solution. The presence of the non-optimal solution reduced experts’ problem solving ability was reduced to about that of players three standard deviations lower in skill level by the presence of the non-optimal solution. Inflexibility of thought induced by prior knowledge (i.e., the blocking effect of the familiar solution) was shown by experts but the more expert they were, the less prone they were to the effect. Inflexibility of experts is both reality and myth. But the greater the level of expertise, the more of a myth it becomes.

Merim Bilalić,Peter McLeoda and Fernand Gobet. 2007.

Comments

[watertank.livejournal.com]

We can define flexibility of thought as the ability to adapt to problems where it is necessary either to use new methods, techniques, knowledge or information, or to modify the existing method of dealing with the problem. In this sense, inflexibility of thought induced by prior experience has been demonstrated experimentally in two situations. In the first, participants can solve the problem with their existing knowledge and do so failing to notice that a superior, but less familiar, solution is available. In the second, a change in the task requires a novel procedure to be generated but experts have difficulty doing this because they cannot suppress the procedures they already possess.

A classic example of the first sort of harmful influence of prior experience on problem solution is the Einstellung (set) effect demonstrated by Luchins (1942) with the water–jug problem. Participants were given five introductory problems that could be solved using the same method. These were termed ‘Einstellung problems’ because the common solution method used for all of them was expected to induce a mental set for solving similar problems. The next two problems, termed ‘critical’, could be solved using the same method but a shorter and simpler solution was also possible. Over 80% of the participants failed to notice the shorter method and continued to use the longer method they had used successfully in the introductory problems. Luchins then presented an ‘extinction’ problem. This could not be solved using the original method but could be solved with the shorter method. Sixty four percent of the participants failed to solve the extinction (1-solution) problem. In comparison, only 5% of a control group who started with the critical (2-solution) problems and did not experience the introductory problems failed to solve the extinction problem (Luchins & Luchins, 1959). Thus, rather than improving their performance, the experimental group’s additional experience of the general problem situation blinded them to a simple solution which was found by almost everyone who had not had the extra experience. Luchins’ striking demonstration that experience can induce inflexibility of thought has been successfully repeated many times in a variety of formats (e.g., Atwood and Polson, 1976, Chen and Mo, 2004, Delaney et al., 2004, Lippman, 1996, Lovett and Anderson, 1996, McKelvie, 1990 and Woltz et al., 2000).

[watertank.livejournal.com]

Einstellung phenomena, in contrast, are notoriously difficult to notice. In the critical, 2-solution problem of an Einstellung task an inefficient method of dealing with the problem is transferred but the method produces a solution. So the person experiencing the effect does not realise that there is any need to look for a better one.