Robert L. McKinley

The Discriminating Power of Items that Measure More than One Dimension.

Determining a correct response to many test items frequently requires more than one ability. This paper describes the characteristics of items of this type by proposing generalizations of the item response theory concepts of discrimination and information. The conceptual framework for these statistics is presented, and the formulas for the statistics are derived for the multidimensional extension of the two-parameter logistic model. Use of the statistics is demonstrated for a form of the ACT Mathematics Usage Test.

Computer Applications to Ability Testing

AbstractPaper-and-pencil tests provide the same set of items for all examinees taking a test and therefore result in some examinees taking items which either are too easy or too difficult for them. Such testing procedures are inefficient and frequently inaccurate, particularly near the limits of the tests range. Tailored testing allows for items to be selected as the process of testing proceeds, with items presented based upon previous responses. Items approximating the ability level of the examinee are selected from a pool of calibrated items, where calibration typically is based upon item difficulty, discrimination, and guessing parameters. This paper describes computer-based procedures for implementing the three-paramater logistic model for item selection in tailored testing and the results of a comparison of a tailored test with a traditional paper-and-pencil test.