Discusses the criteria for an effective diagnotic assessment system. Describes efforts to produce a systematic, flexible assessment system for Foundation-level science. Presents a model for science practical skills, a categorized list of practical skills objectives, and some practical test items. Discusses the relationship between criteria and…

Statistical bias in measurement and ethnic-group bias in testing are discussed, reviewing predictive and construct validity studies. Item bias is reconceptualized to include distance of item content from respondent's experience. Differing values of mean and standard deviation for bias parameter are analyzed in a simulation. References are…

The question bank described which comprises a manual file of questions and a computerized file of question description labels, is designed for use in annual surveys of 11-, 13-, and 15-year-old pupils' science performance. Use of the system for test item construction is also discussed. (JJD)

This paper assesses the reliabilities of the items and dimensions, determines the extent of agreement among raters, identifies the factor structure, and assesses the influence of sex and age of children on raters' perceptions on a short form of the Teacher Temperament Questionnaire. (Author/PN)

Outlines the benefits associated with true-false and multiple-choice tests and sets forth rules for writing effective items for such tests. (FL)

In a two-way analysis of variance and Bonferroni-type follow-up to detect biased items on Forms A and B of the test found that bias-in-item difficulty occurred for the Mexican-American and White groups. The need to examine the issue and reconsider popular conceptualizations of bias in testing is shown. (Author/CM)

The three-parameter logistic model has been used for identifying test items that may be biased. An approach is presented here that is applicable even when one of the subgroups is relatively small. Interactions between item content and group membership, commonly called item bias, may be conceptualized as multidimensionality. (Author/BW)

Two studies evaluated partitioning methods, in order to study how items are perceived by students and to determine how students organize content. Achievement test items in five content areas of educational measurement were used as stimuli to be sorted by groups of students who varied in sophistication. (Author/CTM)

The item bank (Rasch model) has serious limitations for the national assessment of subject performance: it requires agreement about the scope and organization of each school subject and about the difficulty level of each test item; and items tend to focus on factual knowledge, ignoring other desirable learning outcomes. (SJL)

This article examines the potential contribution of latent trait models to the study of intelligence. Nontechnical introductions to both unidimensional and multidimensional latent trait models are given. Multidimensional latent trait models can be used to test alternative multiple component theories of test item processing. (Author/CTM)

Two procedures are presented for detecting violations of the unidimensionality assumption made by latent trait models without requiring factor analysis of inter-item correlation matrices. Both procedures require that departures from unidimensionality be hypothesized beforehand. This is usually possible in achievement tests where several content…

As achievement in basic reading declines, the content validity of tests in special subjects may be reduced because examinees cannot read the test items. This study found that reading ability and item format affected test scores in the area of health science for a sample of high school students. (Author/JKS)

Distinguishing between two types of test questions, constructed-response (essay/completion) and selected-response (multiple-choice/true-false), the author considers both in relation to the specificity of learning outcomes being tested, the use of criterion or norm-referenced measurement, and the formative or summative nature of the evaluation.…

Commenting on the preceding article, pages 8-13 of this journal issue, the author critiques Mason's assessment of the utility of selected-response and constructed-response test questions. (SJL)

A study with 804 undergraduates finds that Minnesota Multiphasic Personality Inventory-2 protocols that contain random responding with an onset after item 370 can be distinguished from protocols with an earlier onset of random responding. Results suggest that protocols with late random responding may be interpretable in terms of basic clinical and…