The effect of choosing a single word response model in the accuracy of estimating parameters for the word two-response

Abstract

Abstract

The study aimed to examine the effect of choosing the two-response word response model (Single-teacher, dual-teacher, and three-teacher), in estimating the singular parameters of the mono-teacher model (difficulty), dual-teacher (difficulty and discrimination), triple-teacher (difficulty distinguishing and guessing) and its accuracy.

            To achieve the goals of the study, the researcher applied a number-reversal test for the Woodcock Johnson III (standard battery) test consisting of 30 two-response items (zero for the wrong answer and one for the correct answer) on a sample of students of the sixth academic preparatory grade of 640 students for both sexes (males and female) from the schools of the province of Baghdad.

            The assumptions of the word response theory were validated (One-dimensional, Local Independence, Single Characteristic Curve, Speed Factor), The results showed that these assumptions were fulfilled. The researcher also tested the suitability of the Chi-square, showing that all vocabulary was appropriate for the models.

            The results showed the following:

1- There were no statistically significant differences in estimating the coefficient of difficulty and singularity due to the model factor (mono-teacher, two-teacher, and three-teacher).

1. 2. There were no statistically significant differences in estimating the discrimination factor and its accuracy due to the approved model (mono-teacher, two-teacher, and three-teacher).
2. 3. The absence of statistically significant differences in estimating the power factor of individuals and their accuracy due to the model factor.
3. 4. Fit of number-reversal test for Woodcock Jones III cognitive ability test (standard battery) for single-parameter model because discrimination coefficients are close and the guess factor is very close to zero.