The application of a marking scheme to psychology projects by student radiographers

A study was initiated to assess the reliability of a marking scheme for research projects on the psychological aspects of patient care. The marking scheme was applied to ten projects by three markers and the Spearman Rank correlation coefficient (rho) was determined for the three pairings of markers. The t-test was applied to the means of scores awarded by the pairs of markers in order to establish the statistical significance of the variation between markers. The Kendall coefficient of concordance was used to measure the relation among the three rankings of scores. The Spearman Rank coefficient and the t-test showed (at p less than 0.05 and p less than 0.10 respectively) that there was a statistically significant difference between one pairing of the scores from the three markers. The Kendall coefficient of concordance was found to be 0.73 with quite a high level of agreement between the three markers, but there were some discrepancies in the rankings. If we take into consideration the individual markers' experiences in assessing such projects, we find that the pair with the most disparity in experience proved to be the pair whose scoring differed significantly. A modified marking scheme was developed as the basis for further research.