The study investigated the degree to which errors of scaling and selection depress the linear relationship and whether the reduction in the magnitude of r differs with the type of error. Results indicated that various scaling errors caused considerable discrepancy in the measurement of underlying relations, but the effect of non-normality was…

Methods for evaluating the consistency of responses to test items were compared. When a researcher is unwilling to make the assumptions of classical test theory, has only a small number of items, or is in a tailored testing context, Cliff's dominance indices may be useful. (Author/CTM)

Two studies focused on aspects of the rating process in performance assessment. The first, which involved 15 raters and about 400 medical students, made the "committee" facet of raters working in groups explicit, and the second, which involved about 200 medical students and four raters, made the "rating-occasion" facet…

Purpose of this study was to examine the effect of using separate, machine scorable answer sheets on the number of marking errors made by third-, fourth-, and sixth-grade students. (Authors)

This study introduces the individual consistency index (ICI), which measures the extent to which patterns of responses to parallel sets of items remain consistent over time. ICI is used as an error diagnostic tool to detect aberrant response patterns resulting from the consistent application of erroneous rules of operation. (Author/PN)

Errors in self-reports of three academic performance measures are analyzed. Empirical errors are shown to depart radically from both no-error and random-error assumptions. Self-reports by females depart farther from the no-error and random-error models for all three performance measures. (Author/BW)

The present study examined the effect of three modes of feedback on the seriousness of error types committed on a post-test. The effect of the feedback made on post-test errors was found to be differential and dependent upon the seriousness of errors committed on the pre-test. (Author/LMO)

Different indices can be used to measure an individual's pattern of responses on an achievement test as usual or consistent with the norm. The relationships among eight of these indices are investigated for a math and reading test given to approximately 2,100 fourth-grade students. (Author/BW)

Approaches proposed for educational diagnostic assessment are reviewed and identified as deficit assessment and error analysis. The development of diagnostic instruments may require a reexamination of existing psychometric models and development of alternative ones. The psychometric and content demands of diagnostic assessment all but require test…

This study examined whether the item response curves from a two-parameter model reflected characteristics of the mathematics items, each of which required unique cognitive tasks. A computer program performed error analysis of test performance. Cognitive subtasks appeared to influence the slopes and difficulties of item response curves. (GDC)

Rater errors (rater severity, halo effect, central tendency, and restriction of range) are described, and criteria are presented for evaluating rating quality based on a many-faceted Rasch (FACETS) model. Ratings of 264 compositions from the Eighth Grade Writing Test in Georgia by 15 raters illustrate the discussion. (SLD)

Advantages and disadvantages of joint maximum likelihood, marginal maximum likelihood, and Bayesian methods of parameter estimation in item response theory are discussed and compared. (Author)

The consistency of student response patterns on a test of language arts was examined in a set of studies as a first step toward designing a computer adaptive test to diagnose errors. (Author/LMO)

A newly introduced approach, rule space, can represent large numbers of erroneous rules of arithmetic operations quantitatively and can predict the likelihood of each erroneous rule. The new model challenges the credibility of the traditional right-or-wrong scoring procedure. (Author/PN)

An index of student patterns of item response, when aggregated to the class level, was shown to identify classes that have a poor match between test content and instructional coverage. The mean caution index for a class can best be interpreted knowing the within-class standard deviation of the index. (Author/LMO)