A decision model designed to determine the reliability and validity of criterion referenced measures (CRMs) is presented. General procedures which pertain to the model are discussed as to: Measures of relationship, Reliability, Validity (content, criterion-oriented, and construct validation), and Item Analysis. The decision model is presented in…

Test bias has traditionally been defined in terms of an outside criterion measure of the performance being predicted by the test. In test construction, where criterion-related validity data are usually not collected until after the test is completed, assessment of bias in the absence of outside criteria had become a vital issue. Here, an unbiased…

Cognitive style research is hindered by a variety of poorly described measures and style classifications and the use of small samples. This study describes and compares three measures of three postulated dimensions of cognitive style: impulsivity/reflectivity; internal/external locus of control; and field dependence/field independence to determine…

A number of different ways of ascertaining whether or not a test measures the same thing in different cultures are examined. Methods range from some that are obvious and simple to those requiring statistical and psychological sophistication. Simpler methods include such things as having candidates "think aloud" and interviewing them about how they…

The steps discussed are (a) stating the purpose of the battery, (b) specifying performance objectives, (c) generating an item pool, (d) item analysis, (e) item selection, (f) determining cut-off scores, and (g) validating the battery. (MP)

Effects that item order and basal and ceiling rules have on test means, variances, and internal consistency estimates for the Peabody Individual Achievement Test mathematics and reading recognition subtests were examined. Items on the math and reading recognition subtests were significantly easier or harder than test placements indicated. (Author)

The effects of option familiarity, length, and technicality on guessing or multiple choice items were investigated in two experiments. Generally, these college undergraduates tended to favor familiar, non-technical, and longer options when guessing on multiple choice tests. (JKS)

Bilevel dimensionality of probability was examined via factor analysis, Rasch latent trait analysis, and classical item analysis. Results suggest that when nonstandardized measures are the criteria for achievement, relying solely on estimates of content validity may lead to erroneous interpretation of test score data. (JKS)

Author-supplied item labels for two common speech discrimination tests were compared with those given spontaneously by 40 children (median age 5:5). Agreement between subjects' and authors' labels was 76.3 percent on the Word Intelligibility by Picture Identification test and 75 percent on the Northwestern University Children's Perception of…

Test Pac, a test scoring and analysis computer program for moderate-sized sample designs using dichotomous response items, performs comprehensive item analyses and multiple reliability estimates. It also performs single-facet generalizability analysis of variance, single-parameter item response theory analyses, test score reporting, and computer…

Spearman correlations were computed between item response theory-derived information functions (IIFs) and four conventional item discrimination indices: phi-coefficient; B-index; phi/phi max; and agreement statistic. Correlations between the phi-coefficient and the IIFs were very high. Data were taken from a real estate licensing test. (Author/GDC)

Standardized methodology was used to help identify item characteristics explaining differential item functioning among Hispanics on Scholastic Aptitude Test (SAT), in two studies with 284,359 and 292,725 Whites, Mexican-Americans, and Puerto Ricans. Results indicate true cognates, or words with a common root in English and Spanish; and content of…

The current study explores the item structure of the Myers-Briggs Type Indicator (MBTI). A factor analysis of responses of a large sample yielded six salient factors, four resembling the four scales of the MBTI. (Author/LMO)

The authors discuss changes that will appear in the 1988 edition of the General Educational Development (GED) test. Changes include (1) a focus on comprehensive, integrated skills; (2) a need for highly developed thinking skills; (3) use of a computer technology theme; (4) writing assessments; and (5) changes in content areas. (CH)