Copyright © 2007-2018 Russ Dewey
Summary: Thinking and Problem Solving
Brain scans show that widespread but specific areas of the brain are involved in different types of thinking. The executive processes, headquartered in the frontal lobes, are important in making plans and guiding activity.
The train of thought can flow smoothly, or it can be derailed by surprises and impediments. Impasses and resets are visible in evoked potentials and also in behavioral indicators such as facial expression.
Ordinarily, focused attention is confined to a limited amount of information. Psychologists studied the problem of selective attention in the 1950s and 1960s.
Broadbent and Cherry proposed early filter theories. Anne Treisman showed that information outside of attention was still being processed for meaning. Ulric Neisser offered a different perspective on attention, depicting it as a goal-directed process, not a filter.
One way to improve intelligence is to develop metacognitive skills. Metacognition is "thinking about thinking" or becoming aware of the nature of one's own thought processes.
Double-checking is a simple form of metacognitive activity. Students who do poorly in school typically show very little self-checking; they are more likely to use one-shot thinking, putting a value on answering a problem quickly rather than thinking it through. When students are taught to slow down and test their own answers, they do better on standardized tests of intelligence.
Cognitive style is the phrase used to describe habitual patterns of thinking. A common distinction is made between analytic and holistic styles of thought. Analytic thinking is typically more valued in the educational system, although holistic thinking may be important for social competence.
All cognitive activity requires some form of problem solving, in the sense that all cognitive activity is goal-oriented. It is aimed at specific productions (thoughts, solutions to problems, percepts, concepts).
The modern era of research on problem solving began with a computer program, the General Problem Solver (GPS). The GPS portrayed a problem as a space. To solve the problem a computer had to find operations (actions) to move from the starting point A to the goal point B.
The main tactic in the General Problem Solver was hill-climbing: taking little steps in the direction of a goal. This can lead to "foothill problem" when stepping toward a goal produces more wasted effort than circling around a problem or impediment to a solution.
Wickelgren listed a variety of approaches to solving problems, based on a course he taught at MIT. His tips such as making sub-goals, working backwards from a goal to discover means of achieving it, and borrowing solutions from different but similar problems.
Another general-purpose technique is to list the constraints or requirements of a problem-solving situation. The ideal solution is one that meets the maximum number of constraints with the fewest drawbacks. Such a solution is likely to be aesthetically satisfying.
An expert is a person with specialized knowledge in a particular area of interest: a domain. A chess expert may be familiar with 10,000 to 100,000 distinct patterns of chess pieces on a chessboard. Expert systems imitate the decision making processes of experts with a list of if-then, condition-action statements.
The computer program SOAR was an attempt to combine various problem solving techniques into one system. The key to SOAR was universal sub-goaling. Each difficulty is treated as a new problem in itself: a sub-goal to be solved. SOAR was not a perfect solution because it required constant intervention by humans, but some form of universal sub-goaling will be part of successful problem solving programs in the future.
Write to Dr. Dewey at email@example.com.