Differences between Scores Assessed by Examiners and Examinees on Objective Structured Clinical Examination

Self-assessment is important for learning. But, the reliability of self-assessment has long been questioned. We investigated this problem in an objective structured clinical examination (OSCE) setting.

Forty third-year nursing students who participated in the OSCE were given the same checklist that was used for assessment by the examiners after they finished the OSCE. Then, the scores that were calculated by the examiners and examinees were compared, measuring the average, correlation, difference in score between examinee and examiner, difference in score according to the examinee's level, and Z-score.

Scores that were made by the examinees were significantly higher than those of the examiners (average 72.34 vs. 64.03), between which there was a loose positive correlation (r=0.34). The correlation coefficient increased (r=0.62) when a difference in score (examinee-examiner) was compared with the examiner score. When examinees were divided into 3 groups according to their scores, the difference tended to become less as the scores of examinee rose. When the Z-score was compared with the level of the examinee, examinees who had higher scores evaluated themselves lower than the examiners, and examinees with lower scores assessed themselves higher than the examiners.

According to comparison of averages or correlation analysis, self-assessment by the examinee appeared to be less reliable. But when data were analyzed using difference in score, examinee levels, and Z-score, a certain tendency developed. This pattern may be due to the heterogeneity of the examinee group. Selection of appropriate analysis methods might be important to estabilish the reliability of self-assessment skills.