Arab Educational Information Network

This study was designed to identify predictors of attrition of graduate students in education, with inclusion of a previously unexamined predictor, provisional admission. The pre-admission predictors considered were undergraduate grade point average, Graduate Record Examination scores, Miller Analogies Test scores, and Georgia Assessment for the Certification of Educators scores. The post-admission predictors considered were number of semesters provisionally admitted, semester of entry, graduate grade point average, and number of course failures. Using binary logistic regression for data obtained from institutional records of 439 students in a graduate education program, no statistically significant pre-admission predictors of graduation were found. The number of semesters provisionally admitted, graduate grade point average, and number of course failures were statistically significant postadmission predictors. In terms of policy implications, limiting the number of terms under which a student may remain provisionally admitted and limiting the number of course failures under which a student may be retained in a program may help reduce student attrition. (Published Abstract)