Figure 1416

Typical log-log plot for the Power Law of Practice.

Practice makes perfect

Common-sense suggests that one way to become an expert is to practice something. Chase and Simon (1973a) estimated that most grand masters had studied for at least 9 to 10 years to reach their level of expertise. The relationship between practice and performance in perceptual-motor skills has been captured by one of the few "laws" in cognitive psychology; the Power Law of Practice (see Figure 14.16). This law states that if the time per trial and number of trials are graphed on log-log co-ordinate axes, then a straight line results (Fitts & Posner, 1967). It is now generally accepted that the power law also holds for purely cognitive skills, so much so that Logan (1988, p. 495) has said that "the power-function speed-up has been accepted as. a law, a benchmark prediction that theories of skill acquisition must make to be serious contenders". Not surprisingly several researchers have suggested a number of mechanisms to explain these effects of practice and the acquisition of expertise: chunking, proceduralisation, and compression (see Chapter 1; Anderson, 1993, 1996; Newell, 1990; Rosenbloom & Newell, 1986). Most of these techniques have been developed in the context of production system models of cognition (Chapter 1).

Practice makes chunks

One proposal is that a form of chunking underlies practice effects (this is a very specific sense of the term, to be distinguished from Miller's, 1956, memory formulation; see Chapter 6). Rosenbloom and Newell (1986, 1987) have argued that when a series of production rules is applied to solve a particular problem, a new rule can be created that does away with the chain of rules (moves) to get to the solution. For example, suppose an problem solver encounters a problem situation, state-a, and needs to reach a goal state, state-f. In solving the problem a chain of rules might be applied: rule 1 changes state-a to state-b, rule 2 changes state-b to state-e, and rule 3 changes state-e to state-f. Stated simply, chunking would create a new rule that contains the relevant conditions that led to the goal state; a new rule that will change state-a to state-f in one step. An immediate implication of chunking is that the problem is solved in one step rather a succession of steps, allowing the time taken to solve the problem to decrease significantly. The idea of a chunk of knowledge has become a standard in many cognitive architectures (e.g., Anderson, 1996; Newell, 1990) and, as we have seen, proves to be a very useful concept in characterising chess expertise.

A schematic diagram of the major components and interlinking processes used in Anderson's (1983, 1993) ACT models. Reprinted by permission of the author.

Learning from problem-solving attempts and instruction: Proceduralisation

Pruning The Elements Of Death

Pruning The Elements Of Death

Learning About Pruning The Elements Of Death In Your Life Can Have Amazing Benefits For Your Life And Success! Discover how you can give yourself new life by cutting off sources of negativity! The concept of positive thinking has interpenetrated our culture. It's a philosophy of faith that doesn't ignore life’s troubles, but instead explains a pragmatic approach to life’s full potentiality.

Get My Free Ebook


Post a comment