أثر عدد بدائل الفقرة و طرق حساب الأخطاء في أنماط الاستجابات وفق تدريج جتمان على قيمتي معامل الاسترجاع و معامل القياسية

Abstract EN

This study aimed at investigating the effect of the number of alternatives item and methods of calculation errors in the response patterns according two ways (Guttman.

Goodenough) on the values of reproducibility and scalability coefficients for testing and item.

To achieve the objectives of the study, it has been applied achievement test in the Mathematics after the confirmation of the psychometric properties, the achievement test included in the final form of (15) items of multiple choice with four Models.

On a sample of (306) students of the scientific first secondary grade students in the Directorate of Education for the first Irbid schools, from the academic year (2014/2015).

The results showed that number of alternatives item, increased the values of reproducibility and scalability coefficients for testing an item according two ways (Guttman.

Goodenough).

The results showed that the increase number of alternatives item increased the value of of reproducibility and scalability coefficients in each of the two methods to calculate the errors in the response patterns and favor to Guttman method.

The results also showed a statistically significant differences at the level of significance (α =0.05) in the reproducibility and scalability coefficients of items due to the method of calculation errors and favor to Guttman method at all items.

As the results showed no a statistically significant differences at the level of significance (α =0.05) in the reproducibility and scalability coefficients of items due to a number of alternatives item and the interaction between a number of alternatives item and the method of calculation errors