Effect of missing values on the matching item in the graded model

Abstract

The study aimed to examine the effect of the percentage of missing values on the matching of the items of the graded model. To achieve this, the experimental approach was followed in conducting the study, as hypothetical data were used to simulate the experimental conditions represented by the percentage of missing values. It has three levels (5%, 10%, and 20%). Moreover, the compensation method is represented by simple regression. The study found that there were no statistically significant differences in the percentage of items matching the graded model according to the missing percentages (5% and 10%). Moreover, there were statistically significant differences in the percentage of items matching the graded model according to the missing percentages (20%). Therefore, we must use the different nonparametric item response theory (IRT) models because of the good information, it provides to researchers about the test and its items without missing items suitable for this purpose and prejudice—the cumulative response of the subjects.