أثر اختيار أنموذج الاستجابة للمفردة (1Pl, 2Pl, 3Pl, 4Pl)‎ و طرائق التقدير في معالم المفردة و الأفراد و دقتها

Abstract EN

The aim of the research is to study the effect of selecting a response model for the item (one parameter, two parameters, three parameters, four parameters) and estimation methods (EAP, MAP and MLP) in estimating the item parameters (difficulty, Discrimination, guess, and lack of interest) and individuals (ability) and its accuracy and detection of the effect of the interaction between These two variables.

To achieve the objectives of the study, the researcher adopted the results of the master's entry test held at the Higher Institute of Languages at the University of Aleppo, consisting of (60) subjects of a double-response objective type (0 or 1) and their response to (1000) advanced and advanced from the theoretical and applied colleges.

After making sure that the assumptions of the response theory of the one-dimensional singularity (using factor analysis, item saturation, double-correlation coefficient), local independence (using chi-square test) and characteristic curve of singularity, the researcher made sure of the suitability of the vocabulary for the two-dimensional models using Jmetrik, the results showed: 1.

There are no statistically significant differences in the proportion of fitting item between the two- parameter, three parameters and four parameters models, while the results showed significant differences in the proportion of fitting item with the one- parameter model.2.

There are no statistically significant differences in the unitary parameters (difficulty, Discrimination and guss) and their accuracy due to the factor of the estimation method, and the existence of statistically significant differences in the value of the item difficulty factor (no significant differences between the one-parameter model and the two-parameter model appeared) and its accuracy, Discrimination, accuracy, and the value of the two-parameter coefficient.

And the effect of the interaction between the estimation method and the model used did not show any indication of differences3.

There are no statistically significant differences in the ability of individuals according to the method of estimation and the model used and the interaction between them, while the results showed that there are statistically significant differences in the accuracy of estimating the ability of individuals according to the model used and the accuracy of the estimate and the interaction between them.

However, the accuracy of estimating the ability did not show significant differences between the two methods.

(EAP and MAP) were not shown, and the three-quadruple model did not show significant differences in the accuracy of the power estimation.