In this article, we introduce a person-fit statistic called the hierarchy consistency index (HCI) to help detect misfitting item response vectors for tests developed and analyzed based on a cognitive model. The HCI ranges from -1.0 to 1.0, with values close to -1.0 indicating that students respond unexpectedly or differently from the responses…

The effects of number of objective-relevant sentences on learning from texts ranging in length from about 800 to 3,000 words were investigated in two related experiments. (Author)

A recursive equation is proposed for computing higher order derivatives of elementary symmetric functions in the Rasch model. A simulation study indicates a small loss in accuracy for the proposed formula compared to Gustafsson's sum algorithm (1980) for computing higher order derivatives when tests contain 60 items or less. (SLD)

Derived shortened versions of the Automatic Thoughts Questionnaire (ATQ) (S. Hollon and P. Kendall, 1980) using samples of 434 and 419 adults. Cross-validation with samples of 163 and 91 adults showed support for the shortened versions. Overall, results suggest that these short forms are useful in measuring cognitions associated with depression.…

In this paper, we apply sequential one-sided confidence interval estimation procedures with beta-protection to adaptive mastery testing. The procedures of fixed-width and fixed proportional accuracy confidence interval estimation can be viewed as extensions of one-sided confidence interval procedures. It can be shown that the adaptive mastery…

Computer administered tests are becoming increasingly prevalent as computer technology becomes more readily available on a large scale. For testing programs that utilize both computer and paper administrations, mode effects are problematic in that they can result in examinee scores that are artificially inflated or deflated. As such, researchers…

The findings of research on short forms of the Wechsler Adult Intelligence Scales-Revised are used to illustrate points about three criteria for evaluating the usefulness of a short form. Results indicate there is little justification for regarding the three criteria as criteria. (Author/BL)

The full version of the Wisconsin Card Sorting Test (WCST) (S. Heaton and others, 1993) was compared to a 64-card version (WCST-64) using normative data and results from 350 neuropsychological patients. The WCST-64 was found useful only for respondents obtaining five or more categories by the end of the first deck. (SLD)

The Wide Range Assessment of Memory and Learning (WRAML) (D. Sheslow and W. Adams, 1990) was given to 51 children. The General Memory Index (GMI) of the WRAML was compared with a short form of the WRAML, the Memory Screening Index (MSI). The MSI was higher than the GMI in 41 of 51 cases. (SLD)

Considers the degree of bias in testlet-based alpha (internal consistency reliability) through hypothetical examples and real test data from four tests of the Iowa Tests of Basic Skills. Presents a simple formula for computing a testlet-based congeneric coefficient. (SLD)

Conducted a Rasch analysis of the psychometric qualities of the Test of Infant Motor Performance (TIMP; G. Girolami and S. Campbell, 1994) for the purpose of reducing the length of the test while maintaining its precision as a measurement device. Using scores from 1,732 tests, the TIMP was reduced to 42 items to form a functional motor scale for…

Several classical approaches and alternative approaches to item selection for computerized adaptive testing (CAT) are reviewed and compared. The study also describes procedures for constrained CAT that may be added to classical item selection approaches to allow them to be used for applied testing. (TJH)

Type I error rates and rejection rates for three-dimensionality assessment procedures were studied with data sets simulated to reflect short tests and small samples. Results show that the G-squared difference test (D. Bock, R. Gibbons, and E. Muraki, 1988) suffered from a severely inflated Type I error rate at all conditions simulated. (SLD)

Presents minimization and maximization models for parallel test construction under constraints. The minimization model constructs weakly and strongly parallel tests of minimum length, while the maximization model constructs weakly and strongly parallel tests with maximum test reliability. (Author/SLD)

DETECT is a nonparametric "full" dimensionality assessment procedure that clusters dichotomously scored items into dimensions and provides a DETECT index of magnitude of multidimensionality. Four factors (test length, sample size, item response theory [IRT] model, and DETECT index) were manipulated in a Monte Carlo study of bias, standard error,…