Determining the Sample Size of the Traditional Test Items Using the Test Item Information Function of the "Rush" Model

Abstract

The study aimed to examine the possibility of using the index of maximum values of information at different levels of measured ability in reducing the items of the "Otis-Lennon" test, relying on a sample of primary, preparatory and secondary education students. The maximum information index associated with the one-parameter "Rush" model was used in analyzing the test data. The study also used the ninth scale to divide the range of the targeted ability measured from the test into nine different ones that vary in terms of the level of measured ability. Then the researcher drew items of different lengths according to the indicator of the maximum value for information. The study found miniatures from the same test, each of which contained a smaller number of items included in the original test. The lowest number of items after the reduction process was (54), which is the shortest form of the test that had the ability to give the same ability ratings as well psychometric characteristics obtained by the original test.