This study explored whether the addition of the items with more advanced math content to the SAT Reasoning Test™ (SAT®) would impact test-taker performance. Two sets of SAT math equating sections were modified to form four subforms each. Different numbers of items with advanced content, taken from the SAT II: Mathematics Level IC Test (Math IC),…

The Student-Problem (S-P) methodology is described using an example of 24 students on a test of 44 items. Information based on the students' test score and the modified caution index is put to diagnostic use. A modification of the S-P methodology is applied to domain-referenced testing. (Author/CM)

Summarizes item review in the development of criterion-referenced tests, including logical item review, which examines the match between instructional intent and the items; empirical item review, which examines response patterns; traditional item review; and instructional sensitivity of test items. Twenty-eight references are listed. (MBR)

Using three samples of children--a heterogeneous group, a hyperactive group, and a replication sample of hyperactive children--evaluted the Stony Brook Scale (SBS). Results indicated the SBS independently assessed hyperactivity and aggression in samples of hyperactive children. (Author/RC)

Two forms of an undergraduate examination were constructed. Tests varied with respect to item truth value (true, false) and method of phrasing (positive, negative). Negatively stated items were more difficult but not more discriminating than positively stated items. False items were not more difficult but were more discriminating than true items.…

Ten item discrimination indices, across a variety of item analysis situations, were compared, based on the validities of tests constructed by using each of the indices to select 40 items from a 100-item pool. Item score data were generated by a computer program and included a simulation of guessing. (Author/CTM)

Reviews information about three major dimensions along which phonics measures vary: (1) group v individual administration, (2) production response (say or write) v recognition response, and (3) the stimuli or materials used to evaluate phonics. (HOD)

Certain kinds of multiple-choice reading comprehension questions may be answered correctly at the higher-than-chance level when they are administered without the accompanying passage. These high risk questions do not necessarily lead to passage dependence invalidity. They threaten but do not prove invalidity. (Author/CP)

Examined cultural bias of the Wechsler Intelligence Scale for Children-Revised (WISC-R) for Anglo-American, Black, and Mexican American children. Minority children responded in the same way as Anglo-American children. No clear pattern to items on the test that were more difficult for minority children appeared. The WISC-R appears to be nonbiased.…

Findings indicated that the judges were not able to determine accurately which items were more difficult for minority students and that there was no significant difference in accuracy between judges of the different ethnic backgrounds. (Author)

Hierarchical cluster analysis is shown to be an effective method for forming scales from sets of items. Comparisons with factor analytic techniques suggest that hierarchical analysis is superior in some respects for scale construction. (Author/JKS)

The author replies to Marx's critique of his essay, "Test Purpose and Item Type," which appears on pages 8-13 of this issue of "Canadian Journal of Education." For Marx's comments, see pages 14-19. (SJL)

A method to measure speededness on tests is presented that reflects the tendency of examinees to guess rapidly on items as time expires. The method models response times with a two-state mixture model, as demonstrated with data from a computer-administered reasoning test taken by 7,218 examinees. (SLD)

Type I error rates for the likelihood ratio test for detecting differential item functioning (DIF) were investigated using Monte Carlo simulations. Type I error rates for the two-parameter model were within theoretically expected values at each alpha level, but those for the three-parameter model were not. (SLD)