مقارنة بين أسلوبي نموذج (راش)‎ اللوغاريتمي و نظرية القياس الكلاسيكية في تحليل و تدريج اختبار لمستويات التفكير الهندسي مبني وفق نظرية (فان هيل)‎

Abstract EN

The present study aims to compare Classical Test Theory (CTT) with Rasch model in Analyzing and calibrating a Criterion Referenced Test (CRT) to measure levels of geometric thinking based on Van Hiele Theory.

The test of geometric thinking levels is a multiple choice test consisted of (28) items.

The test was administered on an exploratory sample and basic one, both of them composed of (681 males and females) students of secondary school.

The results of the test analysis according to the Classical Test Theory show good reliability and validity indices, as well as an acceptable discrimination and difficulty index.

In addition, the results of the calibrated test, after excluding (6) item and (11) person, showed that the test was reliable.

The reliability coefficient of the items is (0.99) and the reliability coefficient of persons is (0.65), and also the items of test covered the difficulty range well, between (-1.61 - 1.46) Logits, and values of the ability of persons between (3.44 - 4.63) Logits, and all the 22 items lies within an acceptable range of MNSQ value (0.7-1.3) and ZSTD value (-0.2,0.2).

The results of the study showed superiority of the "Rasch" model in testing the validity and the reliability of the test and examining the contribution of each item or person in its quality separately