It is widely believed that regression models for binary responses are problematic if we want to compare estimated coefficients from models for different groups or with different explanatory variables. This concern has two forms. The first arises if the binary model is treated as an estimate of a model for an unobserved continuous response and the…

When a psychologist announces a new research finding, it is often based on a rejected null hypothesis. However, if that hypothesis is true, the claim is a false alarm. Many students mistakenly believe that the probability of committing a false alarm equals alpha, the criterion for statistical significance, which is typically set at 5%. Instructors…

Overlap-based measures are increasingly applied in the synthesis of single-subject research. This article considers two criticisms of overlap-based metrics, specifically that they do not measure magnitude of effect and do not adequately correspond with visual analysis. It is argued that these criticisms are based on fundamental misconceptions…

REL Southwest received a request to review the report "Avoidable Losses: High Stakes Accountability and the Dropout Crisis" to assess the soundness of the study methodology and the appropriateness of the conclusions drawn in the report. The review found that conclusions drawn in this study cannot be generalized and are significantly…

Education statistics are rarely neutral; those who collect and analyze them have different purposes. In this article, Bracey discusses several principles of data interpretation to help educators avoid falling into statistical traps. For example, because such reports as A Nation At Risk contain many "selected, spun, distorted, and even manufactured…

Numerous misuses of statistics are described and illustrated, and ways of recognizing and avoiding such misuse are discussed. The following five categories of statistical misuse are identified: a lack of knowledge of the subject matter, the quality of the basic data, the preparation of the study and the report, the statistical methodology, and a…

Offers suggestions for proper procedures for authors to use--and some pitfalls to avoid--when writing studies using factor analysis methods. Discusses distinctions among different methods of analysis, the adequacy of factor structure, and other notes of caution. Encourages authors to ensure that their research is statistically sound. (RJM)

Comments on some methodological limitations of the research base of "The Bell Curve": blind use of the normal distribution (bell curve); avoidance of nonnormal statistical distributions, which are more appropriate for some social and economic characteristics; copious use of percentiles and quintiles, inappropriate with nonnormal…

Using the split-middle methods of trend estimation, evaluates the accuracy of interpretation of single subject data by comparing raters' visual analysis of behavior change with statistical determination of behavior change. Results indicate visual analysis accuracy was less than chance. Rater and student characteristics largely did not affect the…

In a study of the relationship between moral reasoning and teenage drug use, problems arose in an attempt to reduce qualitative data to a quantitative format: (1) making analytic sense of singular and universal responses; (2) the mistaken logical inference that each pattern of judgment should have behavioral indicators; and (3) construction and…

Criticizes data analysis and interpretation in "The Bell Curve:" Herrnstein and Murray do not actually study the "cognitive elite"; do not control for education when examining effects of cognitive ability on occupational outcomes, ignore, cultural diversity within broad ethnic groups (Asian Americans, Latinos), ignore gender…

Too few researchers understand what statistical significance testing does and does not do, and consequently their results are misinterpreted. This Digest explains the concept of statistical significance testing and discusses the meaning of probabilities, the concept of statistical significance, arguments against significance testing,…

Asserts that traditional approaches to multivariate analysis, including standard linear regression techniques, ignore the special character of count data. Explicates three suitable alternatives to standard regression techniques, a simple Poisson regression, a modified Poisson regression, and a negative binomial model. (MJP)

Global figures that reveal the magnitude of the problem of illiteracy in the world disguise great disparities. For example, most of the illiteracy in the world is in developing countries. Therefore, it is preferable, and more appropriate in terms of numerical importance, to concentrate the analysis of illiteracy on developing countries. Such…

The findings of the Cost, Quality, and Child Outcomes (CQO) study in Child Care Centers, the largest and most visible child care research conducted in the 1990s, were widely publicized and used to promote increased spending on initiatives to improve child care quality, the redesign of subsidy systems to provide quality incentives, more stringent…