Playing chess: A study of the transfer of problem-solving skills in students with average and above average intelligence

Philip Jerome Rifner, Purdue University


A study was conducted to determine whether middle-school students who learned general problem-solving skills in one domain could apply them in a different domain. The training task involved learning to play chess, and the transfer task required poetic analysis. The study was conducted in two parts. The first part of the study was a quasi-experiment designed to test whether transfer of training would appear in the form of enhanced performance on twelve dependent variables associated with achievement. The one of primary interest was the rated quality of the subjects' solutions to the transfer task. Others included grades and nine sub-scores and the Total Battery score from the CTBS/4 Achievement Battery. The second investigation was a quantitative-descriptive study conducted to determine which aspects of problem-solving behavior were related to the effects found in the first part. Think-aloud protocols, taken as the subjects solved the transfer problem, were analyzed and coded for problem-solving behaviors. Results indicated several variables of interest: the number of search methods used, the number of goals set, the number of lines considered, the incidence of guessing, the number of unresolved negative evaluations, and the percentage of goals achieved. Both pre- and post-measures were obtained for all variables in both studies, and the results were analyzed using repeated measures analysis of variance. Effect sizes were also calculated for all variables. Results of the quasi-experiment indicated treatment effects only for the transfer task. Results of the quantitative-descriptive study indicated treatment effects for all variables among gifted subjects but only on the number of methods used for students of average ability. The results indicated that inter-domain transfer can be achieved if teaching for transfer is an instructional goal and that transfer occurs more readily and to a greater extent among students with above average ability. Suggestions for future research include exact replications of this study to build its power, which was low, and near replications to extend the generalizability of these findings. Further work in this area may also clarify which teaching methods are most effective in promoting transfer.




Feldhusen, Purdue University.

Subject Area

Educational psychology|Psychology|Experiments

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server