The present paper gives a general introduction to the linear logistic test model (Fischer, 1973), an extension of the Rasch model with linear constraints on item parameters, along with eRm (an R package to estimate different types of Rasch models; Mair, Hatzinger, & Mair, 2014) functions to estimate the model and interpret its parameters. The…

If good measurement depends in part on the estimation of accurate item characteristics, it is essential that test developers become aware of discrepancies that may exist on the item parameters before and after item review. The purpose of this study was to examine the answer changing patterns of students while taking paper-and-pencil multiple…

The full-information maximum likelihood (FIML) method makes it possible to estimate and analyze structural equation models (SEM) even when data are partially missing, enabling incomplete data to contribute to model estimation. The cornerstone of FIML is the missing-at-random (MAR) assumption. In (unidimensional) computerized adaptive testing…

Previous research on the impact of text and formatting changes on test-item performance has produced mixed results. This matter is important because it is generally acknowledged that "any" change to an item requires that it be recalibrated. The present study investigated the effects of seven classes of stylistic changes on item…

The need for increased exam security, improved test formats, more flexible scheduling, better measurement, and more efficient administrative processes has caused testing agencies to consider converting the administration of their exams from paper-and-pencil to computer-based testing (CBT). Many decisions must be made in order to provide an optimal…

Standard unidimensional Rasch models assume that persons with the same ability parameters are comparable. That is, the same interpretation applies to persons with identical ability estimates as regards the underlying mental processes triggered by the test. However, research in cognitive psychology shows that persons at the same trait level may…

For several decades, the "three-parameter logistic model" (3PLM) has been the dominant choice for practitioners in the field of educational measurement for modeling examinees' response data from multiple-choice (MC) items. Past studies, however, have pointed out that the c-parameter of 3PLM should not be interpreted as a guessing parameter. This…

The present study was designed to test for item order effects by measuring four distinct constructs that contribute substantively to anxiety-related psychopathology (i.e., anxiety sensitivity, fear of negative evaluation, injury/illness sensitivity, and intolerance of uncertainty). Participants (n = 999; 71% women) were randomly assigned to…

Computerized classification tests (CCTs) often use sequential item selection which administers items according to maximizing psychometric information at a cut point demarcating passing and failing scores. This paper illustrates why this method of item selection leads to the overexposure of a significant number of items, and the performances of…

The consequences associated with the uses and interpretations of scores for many credentialing testing programs have important implications for a range of stakeholders. Within licensure settings specifically, results from examination programs are often one of the final steps in the process of assessing whether individuals will be allowed to enter…

Computerized classification testing (CCT) is an approach to designing tests with intelligent algorithms, similar to adaptive testing, but specifically designed for the purpose of classifying examinees into categories such as "pass" and "fail." Like adaptive testing for point estimation of ability, the key component is the termination criterion,…

When administering tests across grades, vertical scaling is often employed to place scores from different tests on a common overall scale so that test-takers' progress can be tracked. In order to be able to link the results across grades, however, common items are needed that are included in both test forms. In the literature there seems to be no…

There are relatively few examples of quantitative approaches to quality control in educational assessment and accountability contexts. Among the several techniques that are used in other fields, Shewart charts have been found in a few instances to be applicable in educational settings. This paper describes Shewart charts and gives examples of how…

The first part of this paper describes the use of the robust z[subscript R] statistic to link test forms using the Rasch (or one-parameter logistic) model. The procedure is then extended to the two-parameter and three-parameter logistic and two-parameter partial credit (2PPC) models. A real set of data was used to illustrate the extension. The…

Many testing programs face the practical challenge of having limited resources to conduct comprehensive standard setting studies. Some researchers have suggested that replicating a group's recommended cut score on a full-length test may be possible by using a subset of the items. However, these studies were based on simulated data. This study…