إستخدام نموذج راش في بناء اختبار تحصيلي محكي المرجع لمقرر الإحصاء لطلبة السنة أولى علوم اجتماعية بجامعة زيان عاشور، الجلفة
The study aimed at using the item response theory for constructing a criterion-referenced test in descriptive statistics for 1st year students (social sciences), a common trunk, in the Djelfa University according to research model. The test consisted of (20) MCQ with four alternative items.one of which represents the correct answer. The researcher used the descriptive analytical method because it is the most appropriate for this study. The study sample were (502) students in the Zian Achour University in Djelfa. The most significant results were shown in verifying the assumptions of item response theory in this study, in addition to having 19 items that were inconsistent with the Rasch Model while 1 item were excluded as they found to be inconsistent with the Rasch Model. Furthermore, 18Students were excluded from the sample because they were found inconsistent with the model. The estimates of the difficulty and discrimination coefficients were also acceptable. The analysis also produced a good test that has different forms of validity in addition to (0.99) reliability coefficient of items and (0.87) reliability coefficient of persons. The results also revealed that the maximum value of information was achieved at (e=b) which agrees with Rasch Model and that the test offers the maximum information at medium ability levels while the minimum information was at both lowest and highest ability levels. The results showed that the value of the information function of the test provides the largest amount of information at the average power levels whose value was (3.32) and the standard error of measurement (0.55) corresponding to the middle of the true degree (09), which corresponds to zero scale. It provides the least amount of information at high and low power levels. (Author’s abstract)