بررسی همبستگی نمرات دانشجویان پزشکی دانشگاه علوم پزشکی اردبیل در مقاطع مختلف آموزشی

Introduction: Predictive validity is one of important characteristics of a test; it means that what extent a test predicts scores of criterion test. This type of validity can be obtained through determining correlation between scores of two tests. This study was performed to determine correlation between the scores obtained by medical students in various educational courses, basic sciences and pre-internship comprehensive exams. Methods: In this study, 55 students of Ardabil University of Medical Sciences, accepted within 2004-5 and completed all stages of their training were studied. Information included scores averages in basic sciences, physiopathology, externship, and internship courses and scourges of basic sciences and pre-internship comprehensive exams. Data were analyzed with SPSS software using descriptive statistics (number and percentage) and inferential statistics (Pearson correlation coefficient). Results: The correlations between the averages of each semester from basic science to clinical sciences gradually declined. The highest correlation was between the average of basic sciences and physiopathology courses (r=0.814) and the lowest correlation was between the average of physiopathology and internship courses (r=0.334). On the other hand comprehensive test of basic sciences had the highest correlation with the average of basic sciences course (r=0.746) and gradually, this correlation declined in the next periods. Pre-internship comprehensive exams had the lowest correlation with the average of pre-internship courses (r=0.463). Conclusion: The low correlations between averages ofeach semester from basic science to clinical sciences may cause by different types of training in this courses and the lack of accurate assessment of student performance in the clinical course. The low correlations between scores of comprehensive exams and averages of next courses may cause by using one method of evaluation (multiple choice questions) in these tests and not using new and different methods of evaluations.