Studied three procedures for estimating the standard errors of school passing rates using a generalizability theory model and considered the effects of student sample size. Results show that procedures differ in terms of assumptions about the populations from which students were sampled, and student sample size was found to have a large effect on…

Reference-reliability relates variability due to change and error. It indicates whether some suspected change can be reliably differentiated from random fluctuations. A means by which the process of change can be measured at different points in time is outlined, using empirical data. (TJH)

The estimate-and-subtract method for eliminating threats to validity is described. It is argued that the method is superior to the use of no-difference findings for this purpose. Two ways of improving the no-difference findings are presented. (TJH)

Presents a brief life history of William Sealy Gosset, the "Student" of Student's t-test. Reviews some basic statistics and describes Student's t-test of statistical hypothesis. Contains 11 references. (JRH)

Develops two procedures for estimating individual-level conditional standard errors of measurement for scale scores, assuming tests of dichotomously scored items. Compares the two procedures to a polynomial procedure and a procedure developed by L. Feldt and A. Qualls (1998) using data from the Iowa Tests of Basic Skills. Contains 22 references.…

Studied whether matching on multiple test scores would reduce false-positive error rates compared to matching on a single number-correct score using simulation. False-positive error rates were reduced for most datasets. Findings suggest that assessing the dimensional structure of a test can be important in analysis of differential item functioning…

Investigated incorporating a testlet definition into the estimation of the conditional standard error of measurement (SEM) for tests composed of testlets using five conditional SEM estimation methods. Results from 3,876 tests from the Iowa Tests of Basic Skills and 1,000 simulated responses show that item-based methods provide lower conditional…

Reviews interval estimates of the Pearson, Kendall tau-alpha, and Spearman correlates and proposes an improved standard error for the Spearman correlation. Examines the sample size required to yield a confidence interval having the desired width. Findings show accurate results from a two-stage approximation to the sample size. (SLD)

Derived asymptotic standard errors (SEs) of item response theory equating coefficient estimates using response functions or their transformations. Presents two variations of the item and test response function methods and SEs of their parameter estimates that use logit transformation of the item response functions. Numerical examples show that the…

Examined the precision of change scores in terms of error-tolerance (E/T) ratios for both relative and absolute interpretations of change scores. Used E/T ratios to evaluate the error in estimating the change relative to tolerance for error in a particular context. Illustrates the results with achievement test data. (SLD)

Presents simple and direct demonstrations showing why basic measurement principles require that biases in data created by measurement error be removed. Refutes common objections to the correction for these biases. Also describes substantive psychological processes responsible for some types of measurement error. (SLD)

Submitted the Mathematics Anxiety Rating Scale (MARS) (F. Richardson and R. Suinn, 1972) to a reliability generalization analysis to characterize the variability of measurement error in MARS scores across administrations and identify characteristics predictive of score reliability variations. Results for 67 analyses generally support the internal…

Used logistic regression to model the observation-to-variable ratio required for the standard error scree procedure to identify the number of factors in simulated data correctly. Results demonstrate the ability of logistic regression to simplify summarizing and reporting findings from simulation studies that involve a large number of conditions.…

One version of r-sub(equivalent), calculated from Fisher's exact test p values and recommended for small samples, is considered "a more realistic... [and] a more accurate estimate of the population correlation than... the sample correlation, r-sub(sample)" (R. Rosenthal & D. B. Rubin, 2003, p. 494). Small sample properties of r-sub(sample) and of…

This study evaluated the impact of unequal reliability on test equating methods in the nonequivalent groups with anchor test (NEAT) design. Classical true score-based models were compared in terms of their assumptions about how reliability impacts test scores. These models were related to treatment of population ability differences by different…