Reliability and validity of 4-point and 6-point scales were assessed using a new model-based approach to fit empirical data from 165 graduate students completing an attitude measure. Results suggest that the issue of four- versus six-point scales may depend on the empirical setting. (SLD)

Type I error rates of F. M. Lord's chi square test for differential item functioning were investigated using Monte Carlo simulations with marginal maximum likelihood estimation and marginal Bayesian estimation algorithms. Lord's chi square did not provide useful Type I error control for the three-parameter logistic model at these sample sizes.…

Reports a study to compare the performance of Australian university students on chemistry questions delivered in two different formats: multiple choice and grid. Concluded that students encounter more difficulties with questions presented in the grid format and that grid questions demand a deeper understanding of the topic. (MDH)

College students rated dissimilarity of pairs of common test item formats. A multidimensional scaling model with individual differences fit to data from 111 students suggested that they used 2 dimensions to distinguish among the formats, 1 separating supply from selection items and 1 based on the number of options. (SLD)

The usefulness of global items in predicting weighted-composite evaluations of teaching was evaluated with a sample of 17,183 classes from 105 institutions. Results suggest that, because global items account for a substantial amount of variance, a short evaluation form could capture much of the information needed for summative evaluation. (SLD)

Effects of the following methods for linking metrics on detection of differential item functioning (DIF) were compared: (1) test characteristic curve method (TCC); (2) weighted mean and sigma method; and (3) minimum chi-square method. With large samples, results were essentially the same. With small samples, TCC was most accurate. (SLD)

When college teachers are explicit about specific methods and strategies involved in effective thinking, students are more likely to learn and use these skills. Labeling test questions with the thinking skills required can help students refocus their study methods, resulting in more effective thinking, problem-solving, or decision-making skills.…

Restricted factor analysis shows that the multiple-choice and free-response sections of the Computer Science and Chemistry Advanced Placement examinations (College Board) measure the same proficiencies for the most part. There is a small degree of multidimensionality because of local dependence among free-response items. (SLD)

Equal percentile (EP) and partial credit (PC) equatings for raw scores from performance-based assessments with free-response items are compared through the use of data from the Maryland School Performance Assessment Program. Results suggest that EP and PC methods do not give equivalent results when distributions are markedly skewed. (SLD)

Effects of randomly distributing attitude-measurement items throughout a questionnaire (personality format) versus grouping together items from the same dimension (achievement format) on students' end-of-course evaluations were studied for 376 undergraduates. Advantages demonstrated for the achievement format in terms of statistical results,…

Describes weaknesses of English-as-a-foreign-language (EFL) testing methods and argues that many EFL departments set up their own examinations to assess how effectively they are preparing students for those examinations. It is suggested that this leads to production of test items that are biased against divergent students and that an interview…

Estimating item parameters from a two-parameter normal ogive model is considered using Gibbs sampling to simulate draws from the joint posterior distribution of ability and item parameters. The method gives marginal posterior density estimates for any parameter of interest, as illustrated using data from a 33-item mathematics placement…

Computer simulations were run to measure the relationship between testlet validity and factors of item pool size and testlet length for both adaptive and linearly constructed testlets. Making a testlet adaptive yields only modest increases in aggregate validity because of the peakedness of the typical proficiency distribution. (Author/SLD)

Studies with 220 college students found that multiple-choice test items with 3 items are more difficult than those with 4 items, and items with the none-of-these option are more difficult than those without this option. Neither format manipulation affected item discrimination. Implications for test construction are discussed. (SLD)

A strategy is described for dealing with measurement situations in which certain categories of responses are null, that is, persons do not respond in certain categories to certain items. The method is described for the partial credit model while maintaining the integrity of the original response framework. (SLD)