A general model for the homogeneous case of the continuous response is proposed. The model is an expansion and generalization of the one proposed by the author in 1974, in which the open response situation is dealt with. In this generalized model, the closed response situation is dealt with, and it includes the model for the open response…

This paper outlines a technique for differentially weighting options of a multiple choice test in a fashion that maximizes the item predictive validity. The rule can be applied with different number of categories and the "optimal" number of categories can be determined by significance tests and/or through the R2 criterion. Our theoretical analysis…

The relative difficulty ratio (RDR) is used as a method of representing test difficulty. The RDR is the ratio of a test mean to the ideal mean, the point midway between the perfect score and the mean chance score for the test. The RDR tranformation is a linear scale conversion method but not a linear equating method in the classical sense. The…

For use by those preparing statewide competency testing programs and those developing basic skills examinations at the postsecondary level, this paper provides guidelines for writing items for tests of basic reading skills. Following a discussion of how to choose appropriate items, how to word items, and how to select appropriate reading passages,…

As test standards and research literature in general indicate, definitions of test bias and item bias vary considerably, as do the results of existing methods of identifying biased items. The situation is further complicated by issues of content, context, construct, and criterion. In achievement tests, for example, content validity may impose…

The roles of humor and anxiety in test performance were investigated. Measures of trait anxiety, state anxiety and achievement were obtained on a sample of undergraduate students; the A-Trait and A-State scales of the State-Trait Anxiety Inventory were used. Half of the students received additional humorous items in the achievement test. The…

Two procedures for the display of item analysis statistics are described. One procedure allows for investigation of difficulty; the second plots item difficulty against item discrimination. (Author/JKS)

Sample sentences in grammar books and in tests too frequently deal with cynicism and despair. (DD)

Describes a computer-managed instructional system used to formulate, print, and evaluate true-false questions for testing purposes. The design of the system and its application in medical and nuclear engineering courses in two Spanish institutions of higher learning are detailed. (RAO)

Latent partition analysis is shown to be useful in determining the conceptual homogeneity of an item population. Such item populations are useful for mastery testing. Applications of latent partition analysis in assessing content validity are suggested. (Author/JKS)

This research simulated responses of 75 subjects to 30 items under the Birnbaum model and attempted a fit to the data using the Rasch model. When item discriminations varied from a variance of .05 to .25, there was only a slight increase in lack of fit as the variances increased. (Author/CTM)

Based on a survey of examinations in drama, discusses principles, practices, and issues raised by the findings. (AA)

Comparisons are made between BILOG version 2.2 and LOGIST 5.0 version 2.5 in estimating the item parameters, traits, item characteristic functions, and test characteristic functions for the three-parameter logistic model. Speed and accuracy are reported for a number of 10, 20, and 40-item tests. (Author/GDC)

A Monte Carlo simulation was used to compare the Mahalanobis "D" Squared and the Comrey "Dk" methods of detecting outliers in data sets. Under the conditions investigated, the "D" Squared technique was preferable as an outlier removal statistic. (SLD)

Item response theory equating invariance was examined by simulating vertical equating of two sets of examinee ability data comparing Rasch, three-parameter, and equipercentile equating methods. All three were reasonably invariant, suggesting that multidimensionality is likely to be the cause of lack of invariance found in real data sets. (SLD)