A comparison of two item-scoring procedures and student reactions to them

Two scoring schemes--a partial-credit scoring scheme and a dichotomous scoring approach--were compared to determine whether one could be selected over the other on the basis of empirical data. Means, variances, and reliabilities on alternate measures and student reactions to the two approaches were analyzed. It was found that the mean partial-credit test score was significantly greater than the dichotomous test score. Reliability on the partial-credit measure was higher, but not significantly, than the dichotomously scored measure. Variance of the dichotomously scored measure was greater, but not statistically, than its partial-credit counterpart. Students clearly preferred the partial-credit approach over the dichotomous scheme. It was concluded that the partial-credit scoring system permitted the assessment of a significant amount of partial knowledge. The approach is recommended if rewarding for partial knowledge is an important concern.