To reduce the chance of Heywood cases or nonconvergence in estimating the 2PL or the 3PL model in the marginal maximum likelihood with the expectation-maximization (MML-EM) estimation method, priors for the item slope parameter in the 2PL model or for the pseudo-guessing parameter in the 3PL model can be used and the marginal maximum a posteriori…

The aim of this study was to examine the precision of item parameter estimation in different sample sizes and test lengths under three parameter logistic model (3PL) item response theory (IRT) model, where the trait measured by a test was not normally distributed or had a skewed distribution. In the study, number of categories (1-0), and item…

Mixture Rasch (MixRasch) models conventionally assume normal distributions for latent ability. Previous research has shown that the assumption of normality is often unmet in educational and psychological measurement. When normality is assumed, asymmetry in the actual latent ability distribution has been shown to result in extraction of spurious…

Student misconceptions have been studied for decades from a curricular/instructional perspective and from the assessment/test level perspective. Numerous misconception assessment tools have been developed in order to measure students' misconceptions relative to the correct content. Often, these tools are used to make a variety of educational…

This study analysed polytomous items' psychometric properties according to nonparametric item response theory (NIRT) models. Thus, simulated datasets--three different test lengths (10, 20 and 30 items), three sample distributions (normal, right and left skewed) and three samples sizes (100, 250 and 500)--were generated by conducting 20…

A method is presented for estimating the accuracy and consistency of classifications based on test scores. The reliability of the score is used to estimate effective test length in terms of discrete items. The true-score distribution is estimated by fitting a four-parameter beta model. (SLD)

The Mantel-Haenszel (MH) statistic for identifying differential item functioning (DIF) commonly conditions on the observed test score as a surrogate for conditioning on latent ability. When the comparison group distributions are not completely overlapping (i.e., are incongruent), the observed score represents different levels of latent ability…

The accuracy of estimation procedures in item response theory was studied using Monte Carlo methods and varying sample size, number of subjects, and distribution of ability parameters for: (1) joint maximum likelihood as implemented in the computer program LOGIST; (2) marginal maximum likelihood; and (3) marginal Bayesian procedures as implemented…

Studied the extent to which three appropriateness indexes, Z(sub 3), ECIZ4, and W, are well standardized in a Monte Carlo study. The ECIZ4 most closely approximated a normal distribution, and its skewness and kurtosis were more stable and less affected by test length and item response theory model than the others. (SLD)

This study was designed to compare the accuracy of three commonly used ability estimation procedures under the graded response model. The three methods, maximum likelihood (ML), expected a posteriori (EAP), and maximum a posteriori (MAP), were compared using a recovery study design for two sample sizes, two underlying ability distributions, and…

A simple scheme is proposed for smoothly approximating the ability distribution for relatively long tests, assuming that the item characteristic curves (ICCs) are known or well estimated. The scheme works for a general class of ICCs and is guaranteed to completely recover the theta distribution as the test length increases. The proposed method of…

This study examines a score-difference model for the detection of cheating based on the difference between two scores for an examinee: one based on the appropriate scoring key and another based on an alternative, inappropriate key. It argues that the score-difference method could falsely accuse students as cheaters. (Author/JAZ)

The feasibility of quick norms derived by the procedure described by B. D. Wright and M. H. Stone (1979) was investigated. Norming differences between traditionally calculated means and Rasch "quick" means were examined for simulated data sets of varying sample size, test length, and type of distribution. A 5 by 5 by 2 design with a…

This research examines the effect of increased testing time by comparing the four performance indices of randomly equivalent examinee subpopulations on sections of similar content and difficulty administered at different times on three SAT administrations. A variety of analyses were used in this study and found no evidence that the current SAT…

Effects of test length, sample size, and assumed ability distribution were investigated in a multiple replication Monte Carlo study under the 1-parameter (1P) and 2-parameter (2P) logistic graded model with five score levels. Accuracy and variability of item parameter and ability estimates were examined. Monte Carlo methods were used to evaluate…