Literaturnachweis - Detailanzeige
Autor/inn/en | Olejnik, Stephen F.; Algina, James |
---|---|
Titel | Type I Error Rate and Power of Rank Transform ANOVA When Populations Are Non-Normal and Have Equal Variance. |
Quelle | In: Florida Journal of Educational Research, 27 (1985) 1, S.61-81 (23 Seiten)
PDF als Volltext |
Sprache | englisch |
Dokumenttyp | gedruckt; online; Zeitschriftenaufsatz |
Schlagwörter | Analysis of Variance; Computer Simulation; Error Patterns; Population Distribution; Power (Statistics); Sample Size |
Abstract | The rank transformation approach to analysis of variance (ANOVA) as a solution to the Behrens-Fisher problem was examined. Using simulation methodology, four parameters were manipulated for the two-group design: (1) ratio of population variances; (2) distribution form; (3) sample size; and (4) population mean difference. As a general solution to the group variance inequality problem, the results of this study do not provide sufficient evidence to recommend any single analysis approach. While the rank transform approach was less sensitive to variance inequality than was the parametric ANOVA F-ratio, unacceptably high Type I error rates were obtained when cell frequencies and group variances were inversely related. With equal cell frequencies or when cell frequencies were directly related to group variances, appropriate Type I error rates were obtained. Under these conditions, the Brown-Forsythe procedure for comparing group means provided greater power except when the sampled distribution was leptokurtic. It is contended that before computing hypothesis tests, researchers should first obtain descriptive summary statistics to determine the sample distribution characteristics and to use this information to guide their choice of analysis procedures. Five tables present simulation data. (Author/SLD) |
Erfasst von | ERIC (Education Resources Information Center), Washington, DC |