Two families of analysis methods can be used for differential item functioning (DIF) analysis. One family is DIF analysis based on observed scores, such as the Mantel-Haenszel (MH) and the standardized proportion-correct metric for DIF procedures; the other is analysis based on latent ability, in which the statistic is a measure of departure from…

When using the Rasch model, equating with a nonequivalent groups anchor test design is commonly achieved by adjustment of new form item difficulty using an additive equating constant. Using simulated 5-year data, this report compares 4 approaches to calculating the equating constants and the subsequent impact on equating results. The 4 approaches…

Research has found that the "a"-stratified item selection strategy (STR) for computerized adaptive tests (CATs) may lead to insufficient use of high a items at later stages of the tests and thus to reduced measurement precision. A refined approach, unequal item selection across strata (USTR), effectively improves test precision over the…

An information-correction method for testlet-based tests is introduced. This method takes advantage of both generalizability theory (GT) and item response theory (IRT). The measurement error for the examinee proficiency parameter is often underestimated when a unidimensional conditional-independence IRT model is specified for a testlet dataset. By…

In this study we investigated when an equating conversion line is problematic in terms of gaps and clumps. We suggest using the conditional standard error of measurement (CSEM) to measure the scale scores that are inappropriate in the overall raw-to-scale transformation.

The purpose of this study was to evaluate the combined effects of reduced equating sample size and shortened anchor test length on item response theory (IRT)-based linking and equating results. Data from two independent operational forms of a large-scale testing program were used to establish the baseline results for evaluating the results from…

Equating of tests composed of both discrete and passage-based items using the nonequivalent groups with anchor test (NEAT) design is popular in practice. This study investigated the impact of discrete anchor items and passage-based anchor items on observed score equating via simulation. Results suggested that an anchor with a larger proportion of…

A series of resampling studies was conducted to compare the accuracy of equating in a common item design using four different methods: chained equipercentile equating of smoothed distributions, chained linear equating, chained mean equating, and the circle-arc method. Four operational test forms, each containing more than 100 items, were used for…

Equating under the external anchor design is frequently conducted using scaled scores on the anchor test. However, scaled scores often lead to the unique problem of creating zero frequencies in the score distribution because there may not always be a one-to-one correspondence between raw and scaled scores. For example, raw scores of 17 and 18 may…

The method of maximum-likelihood is typically applied to item response theory (IRT) models when the ability parameter is estimated while conditioning on the true item parameters. In practice, the item parameters are unknown and need to be estimated first from a calibration sample. Lewis (1985) and Zhang and Lu (2007) proposed the expected response…

This study evaluated the impact of unequal reliability on test equating methods in the nonequivalent groups with anchor test (NEAT) design. Classical true score-based models were compared in terms of their assumptions about how reliability impacts test scores. These models were related to treatment of population ability differences by different…

This study examines how closely the kernel equating (KE) method (von Davier, Holland, & Thayer, 2004a) approximates the results of other observed-score equating methods--equipercentile and linear equatings. The study used pseudotests constructed of item responses from a real test to simulate three equating designs: an equivalent groups (EG)…

It is a widely held belief that anchor tests should be miniature versions (i.e., minitests), with respect to content and statistical characteristics of the tests being equated. This paper examines the foundations for this belief. It examines the requirement of statistical representativeness of anchor tests that are content representative. The…

In educational tests, subscores are often generated from a portion of the items in a larger test. Guidelines based on mean-squared error are proposed to indicate whether subscores are worth reporting. Alternatives considered are direct reports of subscores, estimates of subscores based on total score, combined estimates based on subscores and…

This study assessed the factor structure of the LanguEdge™ test and the invariance of its factors across language groups. Confirmatory factor analyses of individual tasks and subsets of items in the four sections of the test, Listening, Reading, Speaking, and Writing, was carried out for Arabic-, Chinese-, and Spanish-speaking test takers. Two…