In this study we examined variations of the nonequivalent groups equating design for tests containing both multiple-choice (MC) and constructed-response (CR) items to determine which design was most effective in producing equivalent scores across the two tests to be equated. Using data from a large-scale exam, this study investigated the use of…

The relationship of multiple-choice and free-response items on the College Board's Advanced Placement Computer Science Examination was studied using confirmatory factor analysis. Results with 2 samples of 1,000 high school students suggested that the most parsimonious fit was achieved using a single factor. Implications for construct validity are…

The consistency of several item bias detection methods was studied across different test administrations of the same items using data from a mathematics test given to approximately 6,600 eighth grade students in all. The Mantel Haenszel and item-response-theory-based sum-of-squares methods were the most consistent. (SLD)

To study whether different forms of the Scholastic Aptitude Test (SAT) used since the mid-1970s varied in their correlations with academic performance criteria, 1975 and 1985 forms were administered to 1,554 and 1,753 high school juniors, respectively. The 1975 form did not have greater validity than the 1985 form. (SLD)

Gender differences in scholastic achievement as a function of measurement method were examined by comparing performance of 739 15-year-old boys and 758 15-year-old girls in Irish high schools on multiple-choice and free-response tests of mathematics, Irish, and English achievement. Method-based gender differences are discussed. (SLD)

Examinees in a regular administration of the quantitative portion of the Graduate Record Examination responded to particular items in a machine-scannable multiple-choice format. Volunteers (n=364) used a computer to answer open-ended counterparts of these items. Scores for both formats demonstrated similar correlational patterns. (SLD)

Figural response items (FRIs) in science were administered to 347 fourth graders, 365 eighth graders, and 322 twelfth graders. Item and test statistics from parallel FRIs and multiple-choice questions illustrate FRIs' more difficult and more discriminating nature. Relevance of guessing to FRIs and diagnostic value of the item type are highlighted.…

Performance of 156 undergraduate and 48 graduate students on a self-adapted test (SFAT)--students choose the difficulty level of their test items--was compared with performance on a computer-adapted test (CAT). Those taking the SFAT obtained higher ability scores and reported lower posttest state anxiety than did CAT takers. (SLD)