Investigation of outliers of evaluation scores among school of health instructors using outlier - determination indices

Introduction: Teacher evaluation, as an important strategyfor improving the quality of education, has been considered byuniversities and leads to a better understanding of the strengthsand weaknesses of education. Analysis of instructors’ scoresis one of the main fields of educational research. Since outliersaffect analysis and interpretation of information processes bothstructurally and conceptually, understanding the methods ofdetecting outliers in collected data can be helpful for scholars,data analysts, and researchers. The present study aimed to presentand compare the available techniques for detecting outliers.Methods: In this cross-sectional study, the statistical populationincluded the evaluation forms of instructors completed bythe students of Shiraz School of Health in the first and secondsemesters of the academic year 2012-2013. All the forms relatedto these years (N=1317) were entered into analysis throughcensus. Then, four methods (Dixon, Gauss, Grubb, and Graphicalmethods) were used for determining outliers. Kappa coefficientwas also used to determine the agreement among the methods.Results: In this study 1317 forms were completed by 203undergraduate and 1114 postgraduate students. The mean scoresgiven by undergraduates and postgraduates were 17.24±3.04 and18.91±1.82, respectively. The results showed that Dixon and Grubbwere the most appropriate methods to determine the outliers ofevaluation scores in small samples, because they had appropriateagreement. On the other hand, NPP and QQ plot were the mostappropriate methods in large samples.Conclusion: The results showed that each of the studied methodscould help us, in some way, determine outliers. Researchers andanalysts who intend to select and use the methods must firstreview the observations with the help of descriptive informationand overview of the distribution. Determination of outliers isimportant in evaluation of instructors, because by determiningthe outliers and removing the data that might have been recordedincorrectly, more accurate and reliable results can be obtained.Keywords: Outliers; Professors; Students; Evaluation