Gender-related differential item functioning (DIF) was examined in a context in which 3,946 middle school students received mathematics instruction focusing on problem solving. Reasons why four tasks on the performance assessment favored female students and two favored male students are discussed. (SLD)

An ideal process is outlined for test item development and the study of item responses to ensure that tests are sound. Qualitative and quantitative methods are used to assess the item-level validity evidence for high-stakes examinations. A checklist for assessment is provided. (SLD)

Nineteen multiple-choice nursing tests containing 673 items were analyzed for test wiseness, irrelevant difficulty in stem or option, linguistic/structural bias, or cultural bias. Twenty-eight types of flaws occurred at least 10 times each. (Contains 28 references.) (SK)

An item-response-theory approach was used to evaluate test and item functioning for the Hare Psychopathy Checklist--Revised (R. D. Hare, 1991) using data from 2,067 North American patients and prisoners. Analyses support the use of the test for diagnosis of psychopathic personality disorder and measures of trait strength. (SLD)

Focuses on the performance levels of the Test of English for Educational Purposes (TEEP) and English Language Testing Service (ELTS) tests. It is concluded that more attention should be paid to the process underlying test performance. (15 references) (GLR)

A method is proposed for the detection of item bias with respect to observed or unobserved subgroups, using a loglinear item response theory model assuming a Rasch model for ability and difficulty. A simulation study was performed with 200 sets of data to check the robustness of the method. (SLD)

The hypothesis that males will perform better on items containing masculine content and girls will score higher on feminine content items was tested with data from a training study intended to improve the ability of 11-year olds to design a controlled experiment. The findings provide weak support for the hypothesis. (Author/YP)

Explores various issues related to empirical validation of computer documentation. Topics discussed include developing validation test items; selecting test users; costs; benefits of testing for the writer; the role of the writer and editors; conducting the validation test; and writing revision specifications. (LRW)

Analysis of the keyed responses of items in three English-as-a-Second-Language reading comprehension tests focused on the form and source of information and on the frequency with which the information appeared in the text. Analysis of the results found that these facets were differentially related to the item score at different proficiency levels.…

The relationship between characteristics of test takers and characteristics of test items was examined in a quasiexperimental study involving 301 high school juniors and sophomores taking a mathematics examination that was consequential to sophomores but not juniors. Results are interpreted in relation to the expectancy value model of motivation.…

Reports on a battery of oral language and early literacy tests, called the SHELL. Describes the tests, presents tasks and scoring system, and provides information about performance by participants in the Home-School Study of Language and Literacy Development. Descriptive, correlational, and predictive analyses based on SHELL-K (kindergarten) and…

This article reports the results of using a three-layer back propagation artificial neural network to predict item difficulty in a reading comprehension test. Three classes of variables were examined: text structure, propositional analysis, and cognitive demand. Results demonstrate that the networks can consistently predict item difficulty. (JL)

Test-taking strategies that examinees may use without reading the passages on which reading comprehension questions are based, similar to those of the new Scholastic Assessment Test, were studied with 350 high school juniors. Strategies most often used involved choosing answers based on consistency and reconstructing main themes from other…

The extent to which items from the Test of English as a Foreign Language and the First Certificate in English function differently for test-takers of equal ability from different native language and curricular backgrounds was investigated. Results suggest a need for methods like logistic regression to examine nonuniform differential item…

Several analyses are presented on the relationships among the Test of Spoken English, Test of Written English, and Test of English as a Foreign Language. The multivariate prediction of each test from the scores on the others is very accurate; variances with two prominent factors may relate to specific cognitive test-taking skills. (eight…