Studied the impact of using five different exposure control algorithms in two sizes of item pool calibrated using the generalized partial credit model. Simulation results show that the a-stratified design, in comparison to a no-exposure control condition, could be used to reduce item exposure and overlap and increase pool use, while degrading…

Compared the nominal and empirical null distributions of the standardized log-likelihood statistic for polytomous items for paper-and-pencil (P&P) and computerized adaptive tests (CATs). Results show that the empirical distribution of the statistic differed from the assumed standard normal distribution for both P&P tests and CATs. Also…

Demonstrates mathematically that if every item in an item pool has an equal possibility to be selected from the pool in a fixed-length computerized adaptive test, the number of overlapping items among an alpha randomly sampled examinees follows the hypergeometric distribution family for alpha greater than or equal to 1. (SLD)

Describes an alternative approach to assessing content validity and content equivalence in terms of item-content structures and content area constructs. Results from applying structural equation modeling to item-response data from two Regents College examinations to verify content constructs suggest the different degrees of inconsistency and bias…

Reviews the literature that focuses on common sources of bias related to the test-generation process, discusses implications of this literature for the business school faculty, and suggests future empirical studies in the areas that pertain to the resolution of these issues. (Author)

Compared is the mass of a dozen eggs with the mass of air in an empty refrigerator. An answer and reasons for the answer are provided. (YP)

A new test critique procedure developed by the National Board of Examiners in Optometry (NBEO) is described. It encourages candidates to identify potentially flawed items during, rather than after, the test. Therefore, candidates retain the opportunity to challenge the test, while the NBEO is able to retain item security. (TJH)

Factors enabling the construction of high-quality tests are clearly defined learning objectives, unambiguous test directions and arrangement, and bias-free statements. (SK)

Polytomous item response theory models and the research that has been conducted to investigate a variety of possible operational procedures for polytomous model-based computerized adaptive testing (CAT) are reviewed, along with studies comparing polytomous CAT systems with competing item response theory models. (SLD)

Suggests methods for developing higher level objective test questions. Taxonomies that define learning outcomes are discussed; and examples for various test types are presented, including multiple correct answers; more complex forms, including classification and multiple true-false; relations and correlates; and interpretive exercises. (LRW)

Methods for detecting item score patterns that are unlikely, given that a parametric item response theory model gives an adequate description of the data or given the responses of other persons in the group, are discussed. The use of person-fit statistics in empirical data analysis is briefly discussed. (SLD)

This article describes a method of incorporating two ability estimates into a single regression analysis so that substantially fewer items exhibit differential item functioning (DIF). Comparative analysis with the Mantel-Haenszel procedure also shows more accurate matching of examinees, preventing identifying multidimensional item impact as DIF.…

Building on previous research, this study examines the elements of the cloze context that affect performance by determining the extent to which item characteristics correlate with item difficulty. (47 references) (LB)

Item response theory (IRT) differential item functioning (DIF) methods used to determine the accuracy of item classification as biased or unbiased were studied. Results from simulations show that the iterative linking and ability scale purification method can be more effective than iterative linking alone primarily by reducing false negatives.…

Two structural laws for intelligence tests are discussed: one law concerns the sign of correlation coefficients and gives conditions under which all correlations between test items will be positive; and one law concerns the relative sizes of the correlation coefficients between intelligence items. A cylindrical structure extends these laws. (SLD)