الأداء التفاضلي لاختبار تحصيلي إلكتروني في مقرر القياس و التقويم لدى طلبة الفرقة الثالثة بكلية التربية بالجامعة باستخدام نموذج التقدير الجزئي

Abstract EN

This research aimed to identify the data to the assumptions of the partial estimation form in the light of grading estimation methods (full-partial), find out how the person's abilities estimation differs from its standard errors in the light of the differences of grading estimation method (full- partial) using partial credit model, find out how the Differential items functioning for the electronic achievement test differs according to the difference of the students' gender(males –females) according to the full method in grading estimation using partial credit model, in addition to find out the best electronic achievement test to judge the efficacy of the information.

The sample consisted of (107) female and male students in education college in technology and applied science university in al Rustaq, in the academic year 2020/2021.

The researchers made the research tool (an electronic achievement test) in the curriculum of measurement and evaluation through blackboard system, and they applied it after checking the item response theory assumptions.

The results (using the statistical analysis program (STATA v.14) were: The electronic achievement test difficulty estimation for grading estimation were between (-2.54: 1.68) using the full method for grade estimation using the partial credit model, The step (0-1) difficulty estimation were between (0.84: 7.38), step (1-2) were between (-7.45: 2.63), step (2-3) were between (-3.29: 2.57), step (3-4) were between (-16.53: 1.08) using the partial method for estimating grade according to the partial estimation form, All Kai values of the electronic achievement test items were not statistically significant except for the step (3-4) in items (1, 2, 3, 8, and 9), The ability estimation using the partial method in grading estimation was higher than the full method in grading estimation, and the standard error of ability estimation using the full method in grading estimation was higher than the standard error of ability estimation using the partial method in grading estimation, There is no statistical differences between male students and female students in the Differential items functioning for the electronic achievement test items using Mantel Hanzel except for item number (1) that indicated a statistical differential performance at the point (0.05) for the male students, There are no statistical differences between male students and female students in the partial performance estimation of the electronic achievement test items using the logistic slope method except for item number (1) which indicated a statistical differential performance at the point of (0.01), as well as There are no statistical differences between male students and female students in the differential performance estimation of the electronic achievement test items using the logistic slope method except for item number (3) which indicated a statistical differential performance at the point of (0.05), The value of the criterion of the information function using Akaike's model is (4107.58) using the full method in grading estimation, and it is (1762.34) using the partial method in grading estimation, which indicated the best model for judging information efficacy is the model depends on the partial method in grading estimation.