Item response theory (IRT) methods are generally used to create score scales for large-scale tests. Research has shown that IRT scales are stable across groups and over time. Most studies have focused on items that are dichotomously scored. Now Rasch and other IRT models are used to create scales for tests that include polytomously scored items.…

A model-testing approach for evaluating the stability of item response theory item parameter estimates (IPEs) in a pretest-posttest design is illustrated. Nineteen items from the Head Start Measures Battery were used. A moderately high degree of stability in the IPEs for 5,510 children assessed on 2 occasions was found. (TJH)

This study examined the consistency of classification for a sample of special needs students on the state-mandated Massachusetts Basic Skills Inventory (BSI). The study sample consisted of 172 special education students (114 males and 58 females) from 15 elementary schools in a large urban school district in Massachusetts, who took the…

The reliability of Mantel-Haenszel (MH) indexes across samples of examinees and sample sizes and their robustness to item context effects were investigated with data for 670 African-American and 5,015 white students from the Second International Mathematics Study. MH procedures can be used to detect differential item functioning. (SLD)