A method of automatically selecting items for inclusion in a test that has constraints on item content and statistical properties was applied to real data. Two tests were assembled by test specialists who assemble such tests on a routine basis. Using the same pool of items and the same

Test forms for two item types were developed that maximized or minimized the differences in average scores between Black and White examinees. The test forms were then administered to random samples of examinees along with conventionally constructed control forms. The results suggest that impact manipulation is possible while maintaining content specifications and average difficulty level. However, the distribution of item difficulties was altered as were the reliability of the experimental sections and the relationships of these sections to other parts of the test.

No abstract

No abstract

No abstract