A simulation study of methods of assessing differential item functioning (DIF) in computer-adaptive tests (CATs) was conducted by Zwick, Thayer, and Wingersky (in press, 1993). Results showed that modified versions of the Mantel-Haenszel and standardization methods work well with CAT data. DIF methods were also investigated for nonadaptive…

In computerized adaptive testing (CAT), new or experimental items are frequently administered alongside operational tests to gather the pretest data needed to replenish and replace item pools. The two basic strategies used to combine pretest and operational items are embedding and appending. Variable-length CATs are preferred because of the…

The relationship of an expert-system-scored constrained free-response item type to multiple-choice and free-response items was studied using data for 614 students on the College Board's Advanced Placement Computer Science (APCS) Examination. Implications for testing and the APCS test are discussed. (SLD)

Findings of studies examining the effects of presentation media on item characteristics are reviewed. The effect of medium of presentation independent of adaptive methodology in computer-assisted testing was studied through tests of 763 Marine trainees. Conditions necessary for score equivalence between item presentation media are described. (SLD)

Discusses factors that should be considered when designing multiple-choice diagnostic tests for use with microcomputer systems. Item content and characteristics are described, presentation of feedback and results are examined, database management functions are discussed, and benefits for students with learning problems and physically disabled…

Effects of reviewing items and altering responses on examinee ability estimates, test precision, test information, decision confidence, and pass/fail status were studied for 376 examinees taking 2 certification tests. Test precision is only slightly affected by review, and average information loss can be recovered by addition of one item. (SLD)

No significant differences in performance on a self-adapted test or anxiety were found for college students (n=218) taking a self-adapted test who selected item difficulty without any prior information, inspected an item before selecting, or answered a typical item and received performance feedback. (SLD)

Computer simulations were run to measure the relationship between testlet validity and factors of item pool size and testlet length for both adaptive and linearly constructed testlets. Making a testlet adaptive yields only modest increases in aggregate validity because of the peakedness of the typical proficiency distribution. (Author/SLD)

Using one simulated and two real data sets, the effects of the systematic variation of the item-selection procedure and the stepsize method on the operating characteristics of computerized adaptive testing (CAT) for instruments with polychotomously scored rating scale items were studied. The six rating scale CAT procedures used performed well.…

Implemented computerized adaptive testing (CAT) to measure students' attitudes toward alcohol. Administered a paper-and-pencil version and a CAT version of an attitudes toward alcohol scale to 113 undergraduates enrolled in health education classes. Findings showed a high correlation between scores from the CAT and the paper-and-pencil versions.…

Exploratory and confirmatory factor analyses were used to explore relationships among existing item types and three new computer-administered item types for the analytical scale of the Graduate Record Examination General Test. Results with 349 students indicate constructs the item types are measuring. (SLD)

Investigates two problems: (1) do learning effects accrue in accuracy or response time when computerized tests are administered in two sessions? and (2) what are the effects of tonal pattern order and contour types on average item difficulty and length of response time for children with different levels of achievement? (DSK)

Evaluated a computer-delivered response type for measuring quantitative skill, the "Generating Examples" (GE) response type, which presents under-determined problems that can have many right answers. Results from 257 graduate students and applicants indicate that GE scores are reasonably reliable, but only moderately related to Graduate…

The purpose of this study was to compare the effects of four item selection rules--(1) Fisher information (F), (2) Fisher information with a posterior distribution (FP), (3) Kullback-Leibler information with a posterior distribution (KP), and (4) completely randomized item selection (RN)--with respect to the precision of trait estimation and the…

Computer-assisted assessment (CAA) has become increasingly important in education in recent years. A variety of computer software systems have been developed to help assess the performance of students at various levels. However, such systems are primarily designed to provide objective assessment of students and analysis of test items, and focus…