Knowledge of the magnitude of effect(s) of an experimental study in science education should be of utmost concern to researchers in the field, but is often not reported. This document describes the concept of "explained variance" in analysis of variance designs and then explains how it can be calculated and reported. Reporting the magnitude of…

THE PURPOSE OF THIS RESEARCH WAS TO STUDY THE EFFECT OF ERROR OF MEASUREMENT UPON THE POWER OF STATISTICAL TESTS. ATTENTION WAS FOCUSED ON THE F-TEST OF THE SINGLE FACTOR ANALYSIS OF VARIANCE. FORMULAS WERE DERIVED TO SHOW THE RELATIONSHIP BETWEEN THE NONCENTRALITY PARAMETERS FOR ANALYSES USING TRUE SCORES AND THOSE USING OBSERVED SCORES. THE…

Subkoviak suggested a technique for estimating both group reliability and the reliability associated with assigning a given individual to a mastery or non-mastery category based on a single test administration. Two assumptions underlie this model. First, it is assumed that had successive test administrations occurred, scores for each individual…

Described and listed herein with concomitant sample input and output is the Fortran IV program which estimates parameters and standard errors of estimate per parameters for parameters estimated through multiple matrix sampling. The specific program is an improved and expanded version of an earlier version. (Author/BJG)

This experiment was designed to assess the ability of item writers to construct truly parallel tests based on a "duplicate-construction experiment" in which Cronbach argues that if the universe description and sampling are ideally refined, the two independently constructed tests will be entirely equivalent, and that within the limits of item…

Livingston's work is a careful analysis of what occurs when one pools two populations with different means, but similar variances and reliability coefficients. However, his work fails to advance reliability theory for the special case of criterion-referenced testing. See ED 042 802 for Livingston's paper. (MS)

A general Computer program is described that will compute asymptotic standard errors and carry out significance tests for an endless variety of (standard and) nonstandard large-sample statistical problems, without requiring the statistician to derive asymptotic standard error formulas. The program assumes that the observations have a multinormal…

A procedure is presented for the analysis of rating data with correlated intrajudge and uncorrelated interjudge measurement errors. Correlations between true scores on different rating dimensions, reliabilities for each judge on each dimension and correlations between intrajudge errors can be estimated given a minimum of three raters and two…

Two criteria are discussed for determining cutting scores on a predictor variable for identifying cases of likely child abuse--utility maximizing and error minimizing. Utility maximizing is the preferable criterion, as it optimizes the balance between the costs of incorrect decisions and the benefits of correct decisions. (Author/JDD)

Ridge estimates (REs) of population beta weights were compared to ordinary least squares (OLS) estimates through computer simulation to evaluate the use of REs in explanatory research. With fixed predictors, there was some question of the consistency of ridge regression, but with random predictors, REs were superior to OLS. (SLD)

An equating design in which two groups of examinees from slightly different populations are administered a different test form with a subset of common items is widely used. This paper presents standard errors and a simulation that verifies the equation for large samples for an equipercentile equating procedure for this design. (Author/BS)

The reliability of criterion-referenced tests (CRTs) used in health program evaluation can be conceptualized in different ways. Formulas are presented for estimating appropriate standard error of measurement (SEM) for CRTs. The SEM can be used in computing confidence intervals for domain score estimates and for a cut-score. (Author/LMO)

Demonstrating good reliability for the difference score in measurement, the results of this study indicate that the difference score is often highly reliable when the correlation between true change and true initial status is nonnegative. In general, when individual differences in true change are appreciable, the difference score shows strong…

Four methods are outlined for estimating or approximating from a single test administration the standard error of measurement of number-right test score at specified ability levels or cutting scores. The methods are illustrated and compared on one set of real test data. (Author)

This study investigated the effect on student scores of using anchor sets that differed in dimensionality in item response theory (IRT) scaled tests. Real data from a mathematics achievement test that had been documented to have dimensions aligned with item format were used. Item responses were available from a representative sample of…