Reviews J. P. Rushton's data in "Race Differences in Behavior: A Review and Evolutionary Analysis" (1988), and suggests that aggregating large cohorts of methodologically weak studies results in misleading conclusions. A reanalysis of Rushton's data shows that cranial size is not a feasible indicator of intelligence and is similar in…

Cross-validated the IMM scale of the Minnesota Multiphasic Personality Inventory-Adolescents (MMPI-A), a measure of ego level, with 151 college students. Means and standard deviations were obtained on IMM scale from the MMPI-A and another MMPI version for males and females. (SLD)

Multiple-choice and true-false versions of Palmore's first Facts on Aging Quiz were completed by 501 college students. Multiple choice reduced the chances of guessing and had less measurement error for average and above-average respondents. (Author/SK)

Provides asymptotic standard errors of the estimates of equated scores from several types of item response theory (IRT) true score equatings. Equating designs considered cover those with internal or external common items and separate or simultaneous estimation. Uses marginal maximum likelihood estimation for the estimation of item parameters. (SLD)

Why the standard error must serve as a standard against which educational gains are measured is discussed from a policy analysis perspective, considering technical and policy levels. An online discussion of the issues in the use of the standard error is transcribed and attached. (SLD)

Evaluated the distribution of person fit within the computerized-adaptive testing (CAT) environment through simulation. Found that, within the CAT environment, these indexes tend not to follow a standard normal distribution. Person fit indexes had means and standard deviations that were quite different from the expected. (SLD)

Content analysis of 364 articles in the Journal of Agricultural Education 1990-1999 indicated that a majority of studies did not mention nonresponse error as a threat to external validity, did not attempt to control for nonresponse error, or did not cite literature on handling it. Protocols to address nonresponse error were proposed. (Contains 24…

Subjective well-being (SWB) is an important indicator of quality of life. SWB can be conceptualized as a momentary state (e.g., mood) as well as a relatively stable trait (e.g., life satisfaction). The validity of self-reported trait aspects of SWB has been questioned by experimental studies showing that SWB judgments seem to be strongly context…

Assessment agencies are increasingly facing pressure on two fronts; first, to increase transparency and openness and second, to improve public confidence. Yet, in relation to one of the central concepts of educational measurement--inherent error--many believe that increased public understanding is incompatible with public confidence: a general…

Getting a "wrong" sign in empirical work is a common phenomenon. Remarkably, econometrics textbooks provide very little information to practitioners on how this problem can arise. The author exposits a long list of ways in which a wrong sign can occur and how it might be corrected.

Ignoring a level can have a substantial impact on the conclusions of a multilevel analysis. For intercept-only models and for balanced data, we derive these effects analytically. For more complex random intercept models or for unbalanced data, a simulation study is performed. Most important effects concern estimates and corresponding standard…

In mean and covariance structure analysis, the chi-square difference test is often applied to evaluate the number of factors, cross-group constraints, and other nested model comparisons. Let model M[a] be the base model within which model M[b] is nested. In practice, this test is commonly used to justify M[b] even when M[a] is misspecified. The…

Type I error rates for PARSCALE's fit statistic were examined. Data were generated to fit the partial credit or graded response model, with test lengths of 10 or 20 items. The ability distribution was simulated to be either normal or uniform. Type I error rates were inflated for the shorter test length and, for the graded-response model, also for…

Two methods, direct maximum likelihood (ML) and the expectation maximization (EM) algorithm, can be used to obtain ML parameter estimates for structural equation models with missing data (MD). Although the 2 methods frequently produce identical parameter estimates, it may be easier to satisfy missing at random assumptions using EM. However, no…

The primary purpose of this study was to examine relative performance of 2 power estimation methods in structural equation modeling. Sample size, alpha level, type of manifest variable, type of specification errors, and size of correlation between constructs were manipulated. Type 1 error rate of the model chi-square test, empirical critical…