Discusses the hypothesis that the difficulty of a multiple-choice comprehension test does not just depend on the difficulty of the text, but rather on the text and the questions. Presents ways to evaluate the difficulty of the text, the questions on it, and the test as a whole. (EKN)

An ignored aspect of standard setting, namely the possibility that Angoff or Nedelsky judges specify inconsistent probabilities (e.g., low probabilities for easy items but large probabilities for hard items) is explored. A latent trait method is proposed to estimate such misspecifications, and an index of consistency is defined. (Author/PN)

Three item-construction principles for multiple-choice tests were studied to determine how they affected test results for high school students: (1) use of incomplete sentence stem; (2) location of blank in the stem; and (3) presence of noncueing material. Differences in item construction had a slight effect on test results. (Authors/CJ)

Multiple matrix sampling of items and examinees was applied to an 18-item rank order instrument administered to a randomly assigned group and compared to the ordering and ranking of all items by control subjects. High correlations between ranks suggest the methodology may viably reduce respondent effort on long rank ordering tasks. (Author/CM)

Investigated cultural bias in 79 items of three verbal tests of the Wechsler Intelligence Scale for Children-Revised (WISC-R). Administered three subtests to 40 Anglo- and 40 Native-American Navajo subjects. Results indicated 15 of 79 items (information, similarities, and vocabulary subtests) were biased against the Navajo sample. (Author)

A study of recurrent faults in multiple-choice items in Britain's Open University's computer-marked tests has led to a procedure for avoiding these faults. A description of the study covers the incidence and sources of faults (obviousness, memorization, unclear instruction, ambiguity, distractors, inter-item effects, and structure) and…

This paper presents a chi square method for assessing bias in test items. In this procedure an unbiased item is defined as an item for which the probability of a correct response is the same for any person of a given ability level regardless of that person's ethnic group. (Author/CTM)

Number right and elimination scores were analyzed on a college level mathematics exam assembled from pretest data. Anxiety measures were administered along with the experimental forms to undergraduates. Results suggest that neither test scores nor attitudes are influenced by item order knowledge thereof, or anxiety level. (Author/GK)

Three case studies are presented that describe variations used to obtain student perceptions of examinations and examination questions. Student perceptions can provide feedback about the quality of the examination. Results are presented and discussed and some examples of test evaluation questions are given. (Author/MLW)

A continuum of item-writing methods is proposed ranging from informal-subjective methods to algorithmic-objective methods. Examples of techniques include objective-based item writing, amplified objectives, item forms, facet design, domain-referenced concept testing, and computerized techniques. (Author/CP)

The effects of two item formats on item difficulty and item discrimination indices for mathematics problem solving multiple-choice tests were investigated. One format required identifying the proper "set-up" for the item; the other format required complete solving of the item. (Author/JKS)

Control groups' scores on the Scholastic Aptitude Test-Verbal were compared to scores of secondary school students who were exposed to their school's special test preparation programs. Special preparation resulted in an overall difference of eight points on the total score, corresponding to one additional correctly answered item. (Author/GDC)

Analysis of variance and subjective rating by curriculum specialists were used to identify biased items on the Iowa Tests of Basic Skills. Results show little agreement between statistical and subjective methods. Test developers should statistically support a reviewer's selection of biased items. (Author/CP)

Reveals that the "Classroom Reading Inventory" is the most passage independent of three informal reading inventories administered to fourth grade students. Shows that the general information, vocabulary meaning, and affective question categories tended to be particularly passage independent. (FL)

Presents remarks on essay testing made by Edward M. White, an account of procedures for developing essay questions for the California State University and Colleges (CSUC) English Equivalency Examination, the 1974 essay scale with accompanying writing samples, and the questions and scale used in the 1975 test. (RL)