Date of Award

2000

Document Type

Thesis

Publisher

Edith Cowan University

Degree Name

Bachelor of Arts (Honours)

Faculty

Faculty of Community Services, Education and Social Sciences.

First Supervisor

Dr Craig Speelman

Abstract

The effect of changing the conceptual context on task performance was examined. The aim was to evaluate whether power functions that describe improvement on old skills during practice can be used to predict further improvement on these skills when they are presented in a novel environment. This research was designed to extend Speelman and Kirsner's (under review) study, which involved testing the assumption made by many skill theories: that performance should continue to improve as if a change in task conditions had no effect. Eighty participants were randomly allocated to one of four distractor conditions: Operand Change (e.g., 2 x 9 = _), Operand Reversal (e.g., 2 x 6 = _), Operation Change (e.g., 6 x _ = 12), and Symbol Change (e.g., 12 + 6 = _), which were designed to vary the amount of conceptual change on a test task. The test items were single-digit multiplication problems from the six-times table (e.g., 6 x 2 = _). Participants repeatedly solved items from the test task during the training phase, and then solved the same task combined with items from a distractor condition during the transfer phase. The results revealed that a change in the conceptual context disrupted immediate transfer performance on the test task in the Operation Change and Symbol Change conditions by slowing response time. A further analysis indicated that performance on the test task remained disrupted for the Operation Change condition, but returned to predicted levels for the Symbol Change condition. It also revealed that performance in the Operand Change condition was disrupted, with performance failing to improve with any further practice. No disruption occurred in the Operand Reversal condition. Accuracy levels remained consistently high and were not influenced by a change in task conditions. These results support Speelman and Kirsner's findings that a change in the presentation context of a task affects response time performance on an established skill, and that power law extrapolations cannot be used to reliably predict transfer performance. These results have implications on skill acquisition and transfer theories that have not accounted for this effect in their transfer predictions.

Download

Share

 
COinS