Practice and transfer in the acquisition of procedural and retrievable knowledge

Raphael Albert Alford, Fordham University

Abstract

Two experiments investigated the acquisition of procedural and retrievable knowledge by examining predictions regarding the role of practice, speed of acquisition, and transfer. In Experiment 1, the participants calculated values to A-a gradient problems in two different practice conditions--varied versus same. In the varied condition, the participants practiced a few times computing values for a large set of problems. After sufficient practice, the participants were expected to become skilled with the computational procedures in calculating the values. In the same condition, the participants solved a small set of problems many times. This practice type was predicted to form associations between the answers and problems and subsequently the associations would facilitate retrieval of the answers to the repeated problems. After practice, each subject calculated the values to a set of novel transfer problems. Results show no significant difference in performance speed for both conditions at the beginning of the task. However, at the end of the task the same condition shows substantially faster reaction times (RTs) than the varied condition. The low level of errors shows that both conditions are very accurate. In the transfer condition, the varied condition shows faster RTs and greater accuracy than the same condition. The practice condition design of Experiment 2 replicated the practice condition design of Experiment 1 with different participants. The participants focused primarily on estimating whether the values were normal or abnormal. Next they received a new set of transfer problems. Half of the participants from each of the practice conditions calculated the values and the remaining half estimated whether the values were normal or abnormal. Analyses of the RTs in the practice condition showed that estimation seems more closely associated with retrievable knowledge. The low level of errors for both conditions was not significantly different. On the transfer calculation task, the varied condition was faster than the same condition. The low level of errors was not significantly different. On the estimation task, even though the same condition was slightly faster at estimation than the varied condition, they were substantially worse at accuracy. Explanations for the findings and recommendations for future studies are discussed.

Subject Area

Educational psychology|Cognitive therapy

Recommended Citation

Alford, Raphael Albert, "Practice and transfer in the acquisition of procedural and retrievable knowledge" (1997). ETD Collection for Fordham University. AAI9808992.
https://research.library.fordham.edu/dissertations/AAI9808992

Share

COinS