Main Article Content

Abstract

This study explored the effect of item difficulty and sample size on the accuracy of equating by using item response theory. This study used simulation data. The equating method was evaluated using an equating criterion (SEE, RMSE). Standard error of equating between the criterion scores and equated scores, and root mean square error of equating (RMSE) were used as measures to compare the method to the criterion equating. The results indicated that the large sample size reduces the standard error of the equating and reduces residuals. The results also showed that different difficulty models tend to produce smaller standard errors and the values of RMSE. The similar difficulty models tend to produce decreasing standard errors and the values of RMSE.

Keywords

Item Response Theory (IRT) equating test.

Article Details