Derives three reliability estimates for the Bayes modal estimate (BME) and the maximum-likelihood estimate (MLE) of theta in computerized adaptive tests (CATs). Computes the three reliability estimates and the true reliabilities of both BME and MLE for seven simulated CATs. Results show the true reliabilities for BME and MLE to be nearly identical…

When examinees are allowed to review responses on an adaptive test, can they "cheat" the adaptive algorithm in order to take an easier test and improve their performance? Theoretically, deliberately answering items incorrectly will lower the examinee ability estimate and easy test items will be administered. If review is then allowed,…

Recent advances in psychometrics and computer technology encourage the development of model-based methods of individualized testing on a microcomputer, where each examinee receives short tests and where the number of pretest items that can be administered is severely restricted. On-line (i.e., data is collected on operational equipment) methods…

Although computerized adaptive tests acquire their efficiency by successively selecting items that provide optimal measurement at each examinee's estimated level of ability, operational testing programs will typically consider additional factors in item selection. In practice, items are generally selected with regard to at least three, often…

This study developed a model to simulate nonmodel-fitting responses in a computerized adaptive testing (CAT) environment, and to examine the effectiveness of the model. The underlying idea was to simulate examinees' test behaviors realistically. This study simulated a situation in which examinees are exposed to or are coached on test items before…

One challenge associated with computerized adaptive testing (CAT) is the maintenance of test and item security while allowing for daily testing. An alternative to continually creating new pools containing an independent set of items would be to consider each CAT pool as a sample of items from a larger collection (referred to as a VAT) rather than…

This paper examines some characteristics of the statistics associated with the sampling distribution of the proficiency level estimate when the Rasch model is used. These characteristics allow the judgment of the meaning to be given to the proficiency level estimate obtained in adaptive testing, and as a consequence, they can illustrate the…

A simulation study explored the effect of population distribution on maximum likelihood estimation (MLE) and expected a posteriori (EAP) estimation in computerized adaptive testing based on the rating scale model of D. Andrich (1978). The choice between EAP and MLE for particular situations is discussed. (SLD)

Basic procedures for performing computerized adaptive testing based on the successive intervals (SI) Rasch model were evaluated. The SI model was applied to simulated and real attitude data sets. Item pools as small as 30 items performed well, and the model appeared practical for Likert-type data. (SLD)

Examined whether providing item review on a computerized adaptive test could be used by examinees to inflate their scores. Two studies involving 139 undergraduates suggest that examinees are not highly proficient at discriminating item difficulty. A simulation study showed the usefulness of a strategy identified by G. Kingsbury (1996) as a way to…

Nonparametric item response models have been developed as alternatives to the relatively inflexible parametric item response models. An open question is whether it is possible and practical to administer computerized adaptive testing with nonparametric models. This paper explores the possibility of computerized adaptive testing when using…

It is unrealistic to suppose that standard item response theory (IRT) models will be appropriate for all new and currently considered computer-based tests. In addition to developing new models, researchers will need to give some attention to the possibility of constructing and analyzing new tests without the aid of strong models. Computerized…

Several person-fit statistics have been proposed to detect item score patterns that do not fit an item response theory model. To classify response patterns as not fitting a model, a distribution of a person-fit statistic is needed. The null distributions of several fit statistics have been investigated using conventionally administered tests, but…

Information provided for computerized adaptive test (CAT) simulees was compared under two conditions on two moderately correlated trait composites, mathematics and reading comprehension. The first condition used information provided by in-scale items alone, while the second condition used information provided by in- and out-of-scale items together…

Computerized adaptive tests (CATs) pose major obstacles to the traditional assessment of differential item functioning (DIF). This paper proposes a modification of the SIBTEST DIF procedure for CATs, called CATSIB. CATSIB matches test takers on estimated ability based on unidimensional item response theory. To control for impact-induced Type I…