This paper discusses the sample size problem and four factors affecting its solution: significance level, statistical power, analysis procedure, and effect size. The interrelationship between these factors is discussed and demonstrated by calculating minimal sample size requirements for a variety of research conditions. (Author)

This study was performed to determine whether hierarchical logistic regression models could reduce the sample size requirements of ordinary (nonhierarchical) logistic regression models. Data from courses with varying class size were randomly partitioned into two halves per course. Grades of students in college algebra courses were obtained from 40…

This Monte Carlo study explored relationships among standard and unstandardized regression coefficients, structural coefficients, multiple R_ squared, and significance level of predictors for a variety of linear regression scenarios. Ten regression models with three predictors were included, and four conditions were varied that were expected to…

Among the most popular techniques used to estimate item response theory (IRT) parameters are those used in the LOGIST and BILOG computer programs. Because of its accuracy with smaller sample sizes or differing test lengths, BILOG has become the standard to which new estimation programs are compared. However, BILOG is still complex and…

This paper is concerned with statistical issues in differential item functioning (DIF). Four subsets of large scale performance assessment data from the Georgia Kindergarten Assessment Program-Revised (N=105,731; N=10,000; N=1,00; and N=100) were analyzed using three DIF detection methods for polytomous items to examine the congruence among the…

CATSIB is a differential item functioning (DIF) assessment methodology for computerized adaptive test (CAT) data. Kernel smoothing (KS) is a technique for nonparametric estimation of item response functions. In this study an attempt has been made to develop a more efficient DIF procedure for CAT data, KS-CATSIB, by combining CATSIB with kernel…

Due to inaccuracies in method, public libraries consistently undermeasure their reference transactions. This article chronicles the process of developing ways to systematize the count. Results of a literature survey, merits of various sample sizes, sample selection, and calculations are all discussed. Appendices include a list of categories of…

A selection model for meta-analysis is proposed that models the selection process and corrects for the consequences of selection by publication on estimates of the mean and variance of the effect parameters. Simulation studies show that the model substantially reduces bias when the model specification is correct. (SLD)

Derived general formulas to determine the sample size requirements for hypothesis testing with desired power and interval estimation with desired precision. Illustrated the approach with the example of a screening test for adolescent attention deficit disorder. (SLD)

The mean and variance of the statistic psi-star, which is useful for measuring agreement between two raters and allows for tests of significance in the case of large samples, are considered. Results may also be applied for the "G" index of J. W. Holley and J. P. Guilford (1964). (SLD)

For Y. Yao's and G. S. James' tests, Type I error rates were estimated for various combinations of the number of variables, sample-size and sample-size-to-variables ratios, and heteroscedasticity. These tests are alternatives to Hotelling's T(sup 2) and are intended for use when variance-covariance matrices are unequal for two independent samples.…

Simplified formulas are proposed for computing the standard errors of equipercentile equating for continuous and discrete test scores. These formulas are easily extended to more complicated equating designs. Results from a study of 719 subjects taking an English test indicated that the formulas work reasonably well for moderate-size samples. (SLD)

In a randomized clinical trial, a researcher can increase statistical power by including a covariate (or pretest) to reduce participants; however, this raises the cost per participant. A simple closed form expression is derived that applied researchers can use to answer the cost/benefit question. (JPS)

Investigates the BILOG computer program's ability to recover known item parameters for different numbers of items, examinees, and variances of the prior distributions of discrimination parameters for the two-parameter logistic item-response theory model. For samples of at least 250 examinees and 15 items, simulation results support using BILOG.…

The alternative proposed by Wilcox (1989) to the James second-order statistic for comparing population means when variances are heterogeneous can sometimes be invalid. The degree to which the procedure is invalid depends on differences in sample size, the expected values of the observations, and population variances. (SLD)