أثر نسب الفقد في بيانات الجذع المشترك في صورة الاختبار للمستويين المنخفض و المرتفع ضمن طريقة تعويض القيم المفقودة في تقديرات القدرة و معالم الفقرات

الملخص EN

The research investigated the effect of percentages of Missingness (5%, 15% and 30%) in the data of anchor items of lower and upper levels of a test form including the method of imputing the missing values in the estimated parameters of ability and items.

Using the 2pl model; responses of (1000) persons for both levels of the test that consists of (80) items; that contains (16) anchor items; were generated using WinGen program by (10) replications.

The assumptions of the unidimensionality and local independence of the generated data had been validated and so on for the reliability.

Package of (Lava-1.6) was used to guarantee the Missingness in each item of the anchor items due to the adopted percentages, then the missing values were imputed using the methods of imputation (ICC, MI, EM).

Means and standard deviations of the estimated parameters of abilities, discrimination and difficulty of the lower and upper levels of the test form had been calculated, and then the repeated measures of ANOVA for them due to the research independent variable; to answer both questions of the research.

The results of both research questions showed; that there were significant differences at (a=0.05) among the means of the estimated parameters of abilities, discrimination and difficulty in the lower and upper levels of the test form due to the independent variable of the research; in favor to the (true, original) situations of estimated parameters of ability, discrimination and difficulty as compared with the rest situations of estimated parameters of ability, discrimination and difficulty.