Effects of item wording on questionnaire reliability and validity were studied, using 280 undergraduate business students who completed a questionnaire comprising 4 item types: (1) regular; (2) polar opposite; (3) negated polar opposite; and (4) negated regular. Implications of results favoring regular and negated regular items are discussed. (SLD)

Cohen and Hyman's response contains several misunderstandings of the original article by Mehrens and Kaminski. One frequently wishes to make inferences to a domain from a test, but teaching a specific performance and testing for that performance does not allow for a domain inference. (SLD)

Hierarchical (adaptive) and linear methods of testlet construction were compared. The performance of 2,080 ninth and tenth graders on a 4-item testlet was used to predict performance on the entire test. The adaptive test was slightly superior as a predictor, but the cost of obtaining that superiority was considerable. (SLD)

Discussion of the use of human markers to mark responses on write-in questions focuses on a study that determined the feasibility of using a computer program to mark write-in responses for the Medical Council of Canada Qualifying Examination. The computer performance was compared with that of physician markers. (seven references) (LRW)

A computerized adaptive test (CAT) based on the nominal response model (NR CAT) was implemented, and the performance of the NR CAT and a CAT based on the three-parameter logistic model was compared. The NR CAT produced trait estimates comparable to those of the three-parameter test. (SLD)

Studied the ability of logical test design (LTD) to predict student performance in reading Roman numerals for 211 sixth graders in Mexico City tested on Roman numeral items varying on LTD-related and non-LTD-related variables. The LTD-related variable item iterativity was found to be the best predictor of item difficulty. (SLD)

The degree to which statistical item selection reduces direction-of-wording effects in balanced affective measures developed from relatively small item pools was investigated with 171 male and 228 female undergraduate and graduate students at 2 U.S. universities. Clearest direction-of-wording effects result from selection of items with high…

This critical evaluation of the Slosson Intelligence Test (SIT) determined that the test items are 30 years old, scores are derived from a nonrepresentative norm group, and scores are not interchangeable with other intelligence measures. The paper concludes that the SIT is unsuited for educational decision-making purposes, including screening,…

Discusses the role of the Rasch model IRT in evaluating two subtests of the Occupational English test and argues for its use in exploring test constructs and in considering the implications of the empirical analysis presented for the validity of communicative language tests involving speaking and writing skills. (39 references) (Author/JL)

Replenishing item pools for on-line ability testing requires innovative and efficient data collection. A method is proposed to collect test item calibration data in an on-line testing environment sequentially using locally D-optimum designs, thereby achieving high Fisher information for the item parameters. (SLD)

This paper describes the use of the Rasch model to help implement computerized administration of the standard and advanced forms of Raven's Progressive Matrices (RPM), to compare relative item difficulties, and to convert scores between the standard and advanced forms. The sample consisted of 95 girls and 95 boys in Australia. (SLD)

Tests experimentally whether stress placement in polysyllabic words differs in Singapore English (SE) and British English (BE), or whether acoustic correlates of stress differ in the two English varieties. Results suggest word-final stress in SE is not result of lexical stress placement, but combination of lengthening of final-syllable words in…

Describes the Rasch calibration of a portion of the Watkins Farnum Performance Scale (J. Watkins and S. Farnum, 1954), a test of instructional music performance, for 218 sixth graders. Results show how Rasch scaling allows item difficulties to be estimated, the test to be administered more efficiently, and diagnostic information to be obtained.…

Individual differences in confidence judgments made by subjects on the accuracy of their answers to psychological test items were studied with 271 Australian college students. Findings suggest that confidence ratings, like the accuracy scores from the tests of human abilities, are stable and reliable measures of between-subjects variability. (SLD)

Used confirmatory factor analysis to examine whether the two-factor solution to the Social Physique Anxiety Scale (E. Hart, M. Leary, and W. Rejeski, 1989) was meaningful. Results for 4 samples of data for college students, high school students, and athletes (n=1,053) from previous studies support the existence of a single substantive factor…