Formulas are provided for estimating statistical power of a test of significance for the difference among means under a variety of conditions. A table for quick power estimates that require no computation for comparing two means in analysis of variance and analysis of covariance is included. (TJH)

Presents solutions for the problem of maximizing the generalizability coefficient under a budget constraint. Shows that the Cauchy-Schwarz inequality can be applied to derive optimal continuous solutions for the number of conditions of each facet. Illustrates the formal similarity between optimization problems in survey sampling and…

The degree of agreement between two raters rating several objects for a single characteristic can be expressed through an association coefficient, such as the Pearson product-moment correlation. How to select an appropriate association coefficient, and the desirable properties and uses of a class of such coefficients--the Euclidean…

The utility of combining confirmatory factor analysis and second-order methods is illustrated in a study of responses of 487 undergraduate and graduate students to the love instrument of C. Hendrick and S. Hendrick. Second-order confirmatory methods allow the researcher to explore complex realities more thoroughly. (SLD)

Improvements are made on previous estimates for the conditional standard error of measurement in prediction, the conditional standard error of estimation (CSEE), and the conditional standard error of prediction (CSEP). Better estimates of how test length affects CSEE and CSEP are derived. (SLD)

Six methods for estimating the standard error of measurement (SEM) at specific score levels are compared by comparing score level SEM estimates from a single test administration to estimates from two test administrations, using Iowa Tests of Basic Skills data for 2,138 examinees. L. S. Feldt's method is preferred. (SLD)

Three potential causes of scale shrinkage (measurement error, restriction of range, and multidimensionality) in item response theory vertical equating are discussed, and a more comprehensive model-based approach to establishing vertical scales is described. Test data from the National Assessment of Educational Progress are used to illustrate the…

Some of the methods originally used to find relationships between reliability and power associated with a single measurement are extended to difference scores. Results, based on explicit power calculations, show that augmenting the reliability of measurement by reducing error score variance can make significance tests of difference more powerful.…

Two articles discuss the controversy about the relationship between reliability and the power of significance tests in response to the discussion of Donald W. Zimmerman, Richard H. Williams, and Bruno D. Zumbo. Lloyd G. Humphreys emphasizes the differences between what statisticians can do and constraints on researchers. Zimmerman, Williams, and…

Some approximate formulas are presented for standard errors of estimated regression coefficients in two-level designs. If the researcher can make a reasonable guess as to parameters occurring in the model, this approximation can be a guide to the choice of sample sizes at either level. (SLD)

Reexamines Ashenfelter and Krueger's estimates of schooling returns (13%), using three additional years of the same twins survey. Finds a 10% return per year of schooling completed. The within-twin regression estimate of schooling's effect in the log wage is smaller than the cross-sectional estimate. (14 references) (MLH)

Summarizes the literature on twins-based estimations of schooling returns. Examines implications of endogenous determination of which twin goes to school longer and of measuring schooling with error. Twin-based estimation may complement other methodologies, but is vulnerable to the same inconsistency that afflicts conventional cross-sectional…

Most tests of congruence in Holland's theory are flawed because they are cross-sectional, use samples of persons already in congruent environments, classify environment casually, fail to relate congruence to relevant aspects of work environments, and have not assessed environment accurately. The centrality of congruence makes the improvement of…

Simulations were used to compare the moderator variable detection capabilities of moderated multiple regression (MMR) and errors-in-variables regression (EIVR). Findings show that EIVR estimates are superior for large samples, but that MMR is better when reliabilities or sample sizes are low. (SLD)

Discusses three types of least squares estimation (generalized, unweighted, and weighted). Results from a Monte Carlo simulation show that, in comparison with other least squares methods, the weighted least squared method generally reduced bias without increasing asymptotic standard errors. (SLD)