Chase and Simon: Perception in Chess
The authors of this experiment set out to perform perception and short-term recall tasks in relation to chess positions for players of different skill levels: M, master chess players, A, class A chess players, and B, beginner players. By asking the participants to recreate chess positions while a) they were looking at a chess position and b) after they had been briefly exposed to a chess position and it was removed from sight, Chase and Simon aimed to observe the differences between “chunking” information in the short-term memory of these players. In their results, they found that the subjects agreed that looking back at the original board many times was easier than trying to retain larger chunks of information in the short-term memory, regardless of skill level. However, the master chess players completed the tasks faster than the class A and beginner players, and seemed to be able to store more information in each “chunk” that they retained.
- · Skill level and learning: Does a master chess player store more, bigger chunks in their short-term memory by analyzing the connection between various smaller piece of information? Can they think more abstractly than novice players?
- · Perception and learning: how does a person perceive and recreate information?
Chi, et.al. Categorization and representation of physics problems by experts and novices
In this paper, the authors studied the differences between how expert and novice thinkers sort, conceptualize, and represent various types of physics problems. Expert thinkers tended to utilize their canonical object frame, or pertinent background information that enhances their understanding of what a question is asking, more commonly than less advanced thinkers. For example, novices categorized sets of problems based on surface structures in the question stems, i.e. what types of key words and diagrams are used in the problem, whereas expert thinkers categorized the problems based on the similarities in the underlying scientific principles that are required to solve each problem. Similar differences were observed when the authors of the experiment asked for descriptions of schemata contents for certain problems; novice descriptions of a physics problem ended with only brief references to underlying scientific principles, whereas these were the first things mentioned by expert physicists.
- · Surface vs. deep learning: expert thinkers seem to rely on underlying concepts of physics and are able to hypothesize solution strategies before they finish reading a question. Novice thinkers often have vaguer solution strategies and are focused more on the actual keywords and visual cues provided in the question stem.
- · Categorizing problems: how thinkers categorize problems can shed light on how they think about problem-solving protocol.
The results of these papers come to similar conclusions regarding the differences in information processing between expert and novice thinkers. Simon and Chase tested the perceptual “chunking” of visual information in chess players of various skill levels, and they found that expert thinkers are able to process information in larger chunks and in a more abstract manner. The Chi paper cites Simon and Chase’s work to incorporate their research on “chunking” knowledge and further elaborating on its prevalence in expert thinkers. Chi’s analysis of problem categorization similarly noted that expert thinkers sort problems based on their underlying scientific principles instead of the keywords provided in the question.