Abstract
This study aimed at investigating the accuracy of estimating the item difficulty parameter of the one parameter logistic model using the pairwise method in light of sample size and test length change. To achieve the study objective, the data of binary responses were generated by five different sample size (50, 100, 250, 500, 1000) examinee, and four different length test (20, 30, 40, 70) item using simra function in the pairwise package in R program by determining the level of the ability and difficulty parameter is between (-3, 3) normally distributed with a mean 0 and a standard deviation 1. To answer the questions of the study, the researcher used R program for estimating the difficulty parameter using the pairwise method, and estimating the ability using the weighted likelihood estimation method. The results of the study revealed statistically significant differences at (α = 0.05) between the standard errors of the estimates of the difficulty parameter due to the sample size for sample size 1000 examinee. The results of the study revealed statistically significant differences at (α = 0.05) between the standard errors of the estimates of the ability parameter due to the sample size and test length and interaction. The study recommended conduct studies which compare between the likelihood method, the Bayes method, and the pairwise method in the accuracy of estimating the parameters of the item.
Recommended Citation
Al-Yassin, Mohammad and Al-Zoubi, Amal
(2022)
"The accuracy of estimating the item difficulty parameter of the one parameter logistic model using the pairwise method in light of sample size and test length change,"
Jordan Journal of Applied Science-Humanities Series: Vol. 31:
Iss.
2, Article 4.
Available at:
https://digitalcommons.aaru.edu.jo/jjoas-h/vol31/iss2/4