It is suggested that error of measurement cannot be routinely incorporated into the "error term" in statistical tests, and that the reliability of test scores does not have the simple relationship to statistical inference that one might expect. (Author/GK)

Discusses classical test theory, including test construction, administration, and use. Covers basic statistical concepts in measurement, reliability, and validity; principles of sound test construction and item analysis; test administration and scoring; procedures for transforming raw test data into scaled scores; and future prospects in test…

Under experimental conditions, Canadian students were subjected to group tests on simple cognitive tasks constructed on the basis of previous theoretical work on perceptual speed. Cross-cultural replications suggest that such tests can be validated within and across cultures. (Author/MJL)

E. Roskam's (1985) conjecture that steeper item characteristic curve (ICC) "a" parameters (slopes) (and higher item total correlations in classical test theory) would be found with more concretely worded test items was tested with results from 925 young adults on the Eysenck Personality Questionnaire (H. Eysenck and S. Eysenck, 1975).…

Classical procedures for calculating the two indices of decision consistency (P and Kappa) for criterion-referenced tests require two testings on each child. Huynh, Peng, and Subkoviak have presented one-testing procedures for these indices. These indices can be estimated without any test administration using Ebel's estimates of the mean, standard…

Conditions under which interaction effects estimated from classical total scores, rather than item response theory trait scores, can be misleading are discussed with reference to analysis of variance (ANOVA). When no interaction effects exist on the true latent variable, spurious interaction effects can be observed from the total score scale. (SLD)

The reliability of a gain is determined by the reliabilities of the components, the correlation between them, and their standard deviations. Reliability is not inherently low, but the components of gains in many investigations make low reliability likely and require caution in the use of gain scores. (SLD)

The critiques by L. Collins and L. Humphreys in this issue illustrate problems with the use of gain scores. Collins' examples show that familiar formulas for the reliability of differences do not reflect the precision of measures of change. Additional examples demonstrate flaws in the conventional approach to reliability. (SLD)

This study empirically examined the behaviors of item and person statistics derived from item response theory and classical test theory, focusing on item and person statistics and using a large-scale statewide assessment. Findings show that the person and item statistics from the two measurement frameworks are quite comparable. (SLD)

Validated data from a developmental mathematics assessment using classical and three-faceted Rasch measurement methods. Analysis of field test data for 289 elementary school students suggested that a unidimensional construct was being measured, as defined by Rasch criteria. Discusses limitations in confirming content-related validity of the…

Classical test theory is founded on the proposition that measurement error, a random latent variable, is a component of the observed score random variable. This article traces the history of the development of classical test theory, beginning in the early 20th century. (SLD)

This article addresses issues in model building and statistical inference in the context of student modeling. The use of probability-based reasoning to explicate hypothesized and empirical relationships and to structure inference in the context of proportional reasoning tasks is discussed. Ideas are illustrated with an example concerning…

For (0, 1) scored multiple-choice tests, a formula giving test reliability as a function of the number of item options is derived, assuming the "knowledge or random guessing model," the parallelism of the new and old tests (apart from the guessing probability), and the assumptions of classical test theory. It is shown that the formula is a more…

The near universality of gender segregation in middle childhood and early adolescence has stimulated extensive research on sex differences in peer relationship processes. Recent reviews of the literature suggest that although some claims of two-cultures theory have clear empirical support, such as strong preference for same-sex peers over…

Purpose: To provide educationalists with an understanding of the key quality issues relating to multiple-choice tests, and a set of guidelines for the quality assurance of such tests. Design/methodology/approach: The discussion of quality issues is structured to reflect the order in which those issues naturally arise. It covers the design of…