Early psychological studies

Early 1900s: Associationists explained problem-solving in terms of finding and strengthening stimulus-response patterns which would deliver solutions (or not): reproductive solutions.In these problems, people were required to reproduce and possibly modify a solution to a previous task. Thorndike (1911), for example, developed the puzzle box, a small cage with many levels and pulleys, that could only be opened by pushing the correct level or pulling the right string attached to a pulley. Thorndike observed how hungry cats learnt to escape such boxes, and argued from this behaviour that problem solving involved the application of "habit families". A habit family was a set of responses available to an individual in a given situation. Habits within a family were argued initially to be tried randomly, but learning was held to lead to the reorganisation of habit families, with successful responses being strengthened or promoted (and hence becoming more likely with practice) and unsuccessful responses being weakened or demoted (and hence becoming less likely with practice)

Kohler (1925) examined the problem solving abilities of apes. When faced with novel problems (e.g., a number of crates, a stick and a bunch of bananas, apes sat in thought before carrying out a complete solution - building a stack of crates, then using a stick). Kohler argued that this could not be explained by the direct application of habit families.

Associationists maintained the view that problem solving was the application of habit families by suggesting that habit families could be applied covertly. According to this view Kohler's apes assembled random sequences of responses in their heads until a sequence was found that solved the problem. This sequence was then applied in full.

1940s

Focused on productive problems - problems whose solutions involved more than simply reproducing an earlier response. Insight problems 

Dunker's (1945) X-ray problem The problem involved finding a way of using X-rays to destroy a tumour without damaging the surrounding healthy tissue. A strong dose of X-rays would destroy the tumour, but this dose would also kill surrounding tissue. The solution is to project several weak X-ray sources from different directions so they all intersect at the tumour site, producing a concentration of X-rays at the tumour but not at the surrounding tissue.

Gestalt psychologists viewed problem solving as a process of imposing an appropriate structure on the problem solving task. It was argued that once the appropriate structure was found the solution would be straightforward. As such, Gestalt psychologists were particularly interested in what prevented successful problem solving. One of the most robust findings was that prior problem solving performance could create in people a "mental set" in which they were predisposed towards a particular solution strategy, even if that strategy was inappropriate for the problem under consideration.

Neither approach had much place for cognitive activity.

Neither the associationist nor the Gestalt view of problem solving provides an adequate account of the complexities of human problem solving. This much is obvious from the fact the views focus on different types of problems (reproductive vs. productive)

The Associationist school, with its concentration on stimulus-response associations, cannot account for the effects of knowledge (both general strategic knowledge and task-specific knowledge) on problem solving

The Gestalt school, with its emphasis on structure and its appeal to incompletely specified processes such as illumination, fails to provide an adequately concrete account of the processes between stimulus (the presentation of a problem) and response (problem solving behaviour

CHANGED BY WORK OF HERB SIMON IN 1970S

Associationist studies

Gestalt psychologists

Cognitive activity