A technique to estimate variances of weighted estimates in program evaluation is discussed. BRR estimates of variance are constructed from orthogonally weighted subsamples and can provide better estimates from complex samples than can other methods. Use of BRR is illustrated in a government evaluation of a food service program. (SLD)

Two forms, pure-rank and mixed-rank, of a nonparametric, general, linear model-based statistic that can be used to test several hypotheses are presented. A Monte Carlo study was used to investigate the distributional properties of these forms, and their use is discussed. (SLD)

Estimated Type I error rates and power are reported for six tests of variance. Normal and non-normal distributions of a two-group design were investigated. Tests developed by M. B. Brown and A. B. Forsythe and by R. G. O'Brien were considered most useful. Implications are discussed. (SLD)

An effective method of teaching statistical ideas is to show students how to use a computer package to analyze survey data that they have collected. This paper discusses the advantages of this approach and illustrates it by considering instruction on summarizing distributions of individual variables. The importance of stating results and…

Describes a study that examined Top 40 singles chart data to determine whether the frequency distribution of artist productivity fit either of two laws of scatter (Lotka Law of Scientific Productivity or Bradford Law of Scatter). Possible reasons for the lack of statistical significance found between the theoretical and observed distributions are…

Tells how to conduct a meta-analysis study, and refers readers to information about refinements of the techniques. A simple meta-analysis is presented. Describes three fundamental stages to conducting a meta-analysis: (1) literature review; (2) calculation of effect sizes; and (3) analysis of data. (Author/NB)

Data from a study of 48 hospitalized Mexican American diabetics suggests that concurrent enrollment in a health education program, by the patient and a significant other, produces greater compliance with therapeutic regimes than when significant others are not involved. (IAH)

Since research conclusions derived from inappropriate statistical methodologies can lack accuracy, the article describes a technique appropriate to the heterogeneity of special needs groups in which subjects are assigned to fairly homogeneous subgroups. The technique is related to research in reading with this population. (DB)

Presents an experiment to compare the effectiveness of 95% ethanol to 20% bleach as disinfectants using t-tests for the statistical analysis of the data. Reports that bleach is a better disinfectant. Discusses the statistical and practical significance of the results. (JRH)

This study determined through psychophysical comparison of scaling data that speech naturalness judgments of stutterers and nonstutterers from audiovisual recordings form a metathetic (or qualitative) rather than prothetic (or quantitative) continuum. Both direct magnitude estimation and equal-appearing interval scaling were valid, but interval…

A maximum test in which the test statistic is the more extreme of the Brown-Forsythe and in which O'Brien's test statistics are developed, with estimated Type I error rates and power for all three tests. For study conditions, Type I error rates for the maximum test are near the nominal level. (SLD)

Comparisons of subjects are specifically objective if they do not depend on the items involved. Such comparisons are not restricted to the one-parameter logistic latent trait model but may also be defined within ordinal independence models and even within the two-parameter logistic model. (Author)

Reviews three functions of the "Scientist" software package useful for the social sciences: nonlinear curve fitting, parameter estimation, and data/regression plotting. Social scientists are likely to find limitations and unfamiliar procedures in "Scientist". Its value lies in its visual presentation of data and regression curves and the…

Notes that presence of heterogeneous regression slopes in an analysis of covariance design is equivalent to presence of interaction in randomized-blocks design. Describes modification of the usual graphical representation of heterogeneous regressions that can aid in interpreting significant regions for regression surfaces. (NB)

Notes that family researchers can examine extent of self-selection bias in samples by using range of "plausibly correlated characteristics" such as marriage and divorce public records. Provides extensive case example of analyses and discusses compensation techniques of weighting and hazard rate models. (Author/NB)