Date of Award

2010

Degree Type

Thesis

Degree Name

Bachelor of Arts Honours

School

School of Psychology and Social Sciences

Faculty

Faculty of Computing, Health and Science

First Advisor

Dr Craig Speelman

Abstract

Research into the effect of automaticity on skill transfer has resulted in conflicting conclusions about how automatic processes act on the transferability of skill. The research in this study was designed to investigate the existence and nature of the relationship between automaticity in skill acquisition and the ability to transfer that skill to a different task. Using a quantitative research design, a simple counting exercise was used to train participants in a skill, with the amount of training manipulated between groups. Accuracy rates and reaction times were recorded and analysed to determine the variance within and between the groups between initial training, final training, and transfer blocks to gauge the degree of skill transfer as a function of the amount of practice (degree of automaticity). Theories of ACT (Anderson, 1981, 1982, 1983) and Instance theory (Logan, 1988) of skill acquisition are detailed and applied to outcomes to explain characteristics of the underlying mechanisms of acquiring skill and the ways they account for improvements toward automatic performance. Furthermore, varied research and views on the affects of automaticity on skill transfer are outlined and applied to elucidate the outcomes in the analyses of results. Outcomes indicated that performance improved with increased training. This was demonstrated both over trials within the groups, and by comparisons between the groups in their final training blocks. Linear regression analysis of the data was conducted to observe changes in performance as a function of the number of stars that appeared in the stimuli. These too showed greater levels of automaticity were approached with extended training. The participants who received the most training showed less variation in performance despite the numerosity of stars than did those who received less training by the end of the training phase. Finally, correlation analysis between slope (m) values and reaction time differences between final training and transfer blocks indicated that those participants who received the greatest amount of training also experienced the greatest amount of disruption to performance when presented with the initial transfer task. The findings of the study are discussed in relation to skill acquisition and previous observations of the effects of automatic performance on transfer. It is concluded that the results indicate it is possible that varied degrees of automaticity could be used to gauge skill transfer.

Share

 
COinS