The effectiveness of four different imputation procedures (mean, conditional mean, hot deck, and regression) was compared in a two-variable prediction system using a sample of 18,869 college students. Results show that the grand mean procedure is not appropriate for handling missing data. Conditions under which either the conditional mean or…

The bias and relative efficiency of two alternative estimators of optimal weights in meta-analysis were assessed through Monte Carlo simulation, defining the standardized mean differences as the effect-size index. The estimator proposed by L. Hedges and I. Olkin (1985) was more efficient than that of J. Hunter and F. Schmidt (1990). (SLD)

Two bootstrap tests are proposed to test the independence hypothesis in a two-way cross table. Monte Carlo studies are used to compare the traditional asymptotic test with these bootstrap methods, and the bootstrap methods are found superior in two ways: control of Type I error and statistical power. (SLD)

The validity of a higher order factorial structure of a Bulgarian version of the Beck Depression Inventory (BDI) (A. Beck et al, 1961) for nonclinical adolescents was studied with 3 samples totaling 691. Findings yielded strong support for the hypothesized second-order factor, adding to a growing cross-cultural agglomerate of construct validity…

The National Center for Student Aspirations developed the Student Aspirations Survey to address issues of student aspirations at the secondary level. The developed instrument, 98 items with 13 scales, was administered to 1,674 students from rural and suburban high schools. Results provide evidence of internal consistency, although the factor…

The dimensionality of students' attitudes toward regular classroom activities and enrichment clusters were studied with 379 urban elementary school students using two instruments developed for the study. Analyses allowed examination of the construct validity of the scores from both measures while indicating that derived factors were not the same…

The structure of subscores obtained through streamlined scoring of 334 adults' responses to Figural Forms A and B of the Torrance Tests of Creative Thinking (P. Torrance, 1966) was analyzed. Principal components analyses indicated that one general creativity factor adequately represented subscores of both forms. However, the five subscores of each…

This study tested the specificity-generality hypothesis regarding moderation of aptitude test validity by job ability requirement differences using 24,482 Air Force enlistees in 37 jobs. Moderating effects due to job differences were not found, and job ability differences did not moderate the relationship between the amount of "g" measured by a…

This study assessed the accuracy of parallel analysis, a technique in which observed eigenvalues are compared to eigenvalues from simulated data when no real factors are present. Three studies with manipulated sizes of real factors and sample sizes illustrate the importance of modeling the data more closely when parallel analysis is used. (SLD)

A simulation study was conducted to investigate the application of expected a posteriori (EAP) trait estimation in computerized adaptive tests (CAT) based on the partial credit model and compare it with maximum likelihood estimation (MLE). Results show the conditions under which EAP and MLE provide relatively accurate estimation in CAT. (SLD)

The relationship between class size and student evaluation of university teaching quality was analyzed. Data from 2915 university classrooms were collected in classes ranging from 1 to 234 students. Results indicate that there is a weak relationship between class size and student ratings of teaching quality, and that this is a nonlinear…

The effect of removing nonfunctioning items from multiple-choice tests was studied by examining change in difficulty, discrimination, and dimensionality. Results provide additional support for the benefits of eliminating nonfunctioning options, such as enhanced score reliability, reduced testing time, potential for broader domain sampling, and…

The reliability and construct validity of scores on the Personal Preferences Self-Description Questionnaire (PPSDQ) (B. Thompson) were studied. Classical psychometric analysis of data from 641 participants (item and alpha analyses and both oblique and orthogonal exploratory factor analyses) indicate that the PPSDQ has reasonable properties. (SLD)

The factor structure of three commonly used measures of mathematics anxiety was examined using a sample of 323 undergraduates in an algebra class. Six oblique factors were identified, but results did not support the existence of a single common dimension of mathematics anxiety as represented by the six factors. (SLD)

Three analytic approaches were used in a framework of classical test theory to examine the construct validity of a mathematics assessment of 16 constructed response items. Results from 280 elementary school students across four age groups suggest a developmental structure of tasks and subdomains that was generally consistent with the test's…