The psychometric properties of the preclinical and clinical phase testing: A sample of the Ege University

The aim of this study is to investigate the psychometric properties of achievement tests applied to 3rd and 4th grades of Ege University Faculty of Medicine. The sample of the pre-clinical education period of the study group consisted of students who were 3rd grade in the 2012-2013 academic year and passed the achievement test. The same students, 4th grade in the 2013-2014 academic year The data were analyzed by using WINSTEPS and SPSS Statistics 23 programs. In the Rasch model analysis, it was concluded that the test showed good compliance with Rasch model and was reliable by looking at the substance and responder values in the 3rd and 4th class tests. The data of the tests were found to be compatible with Dichotomus Rasch Model and it was observed that the test participants showed a balanced distribution with their skill levels. In the tests, one-dimensionality and local independence assumption of the Rasch model were also examined and it was concluded that it also fulfilled these assumptions. Differential Item Functioning levels (DIF) varying according to gender were examined and no significant difference was found. However, it was concluded that the third grade achievement test performances in the pre-clinical education period predicted the fourth grade achievement test performances in the study group by regression analysis. Keywords: Rasch model, Medical Education, Psychometry