Title:
Effect sizes for research : univariate and multivariate applications
Author:
Grissom, Robert J.
ISBN:
9780415877688
9780415877695
Personal Author:
Edition:
2nd ed.
Publication Information:
New York : Routledge, c2012.
Physical Description:
xvii, 434 p. : ill. ; 23 cm.
General Note:
Previous ed.: London: Lawrence Erlbaum, 2005.
Formerly CIP.
Contents:
Machine generated contents note: ch. 1 Introduction -- Introduction -- Null-Hypothesis Significance Testing -- Statistically Signifying and Practical Significance -- Definition, Characteristics, and Uses of Effect Sizes -- Some Factors Influencing Effect Sizes -- Controversy About Null-Hypothesis Significance Testing -- Purpose of This Book -- Power Analysis -- Replication and Meta-Analysis -- Assumptions of Test Statistics and Effect Sizes -- Violations of Assumptions Suggested by Real Data -- Exploring the Data for a Possible Effect of a Treatment on Variability -- Worked Examples of Measures of Variability -- Summary -- Questions -- ch. 2 Confidence Intervals for Comparing the Averages of Two Groups -- Introduction -- Ratio-of-Means Effect Size -- Background -- Confidence Intervals for μa -- μb: Independent Groups -- Frequentist and Bayesian Perspectives -- Equivalence Testing, Non-Inferiority, and Superiority -- Worked Example for Independent Groups --
Contents note continued: Further Discussions and Methods -- Solutions to Violations of Assumptions: Welch's Approximate Method -- Worked Example of the Welch Method -- Yuen's Confidence Interval for the Difference Between Two Trimmed Means -- Other Methods for Independent Groups -- Criteria for Methods for Constructing a Confidence Interval -- Dependent Groups -- Summary -- Questions -- ch. 3 The Standardized Difference Between Means -- Introduction -- Standardized Difference Between Treatment and Comparison Means Assuming Normality -- Uses and Limitations of a Standardized Difference -- Equal or Unequal Variances -- Outliers and Standardized-Difference Effect Sizes -- Tentative Recommendations -- Additional Standardized-Difference Effect Sizes When There Are Outliers -- Confidence Intervals for Standardized-Difference Effect Sizes -- Counternull Effect Size -- Extreme Groups -- Percent of Maximum Possible Score -- Dependent Groups --
Contents note continued: Effect Sizes for Pretest--Posttest Control-Group Designs -- Summary -- Questions -- ch. 4 Correlational Effect Sizes and Related Topics -- Introduction -- Dichotomizing and Correlation -- Point-Biserial Correlation -- Unequal Base Rates in Nonexperimental Research -- Correcting for Bias -- Confidence Intervals for rpop -- Null--Counternull Interval for rpop -- Assumptions of Correlation and Point-Biserial Correlation -- Unequal Sample Sizes in Experimental Research -- Unreliability -- Adjusting Effect Sizes for Unreliability -- Restricted Range -- Scale Coarseness -- Small, Medium, and Large Effect Size Values -- Binomial Effect Size Display -- Coefficient of Determination -- Criticisms of the Coefficient of Determination -- Slopes as Effect Sizes -- Effect Sizes for Mediating and Moderating Variables -- Summary -- Questions -- ch. 5 Effect Size Measures That Go Beyond Comparing Two Averages -- Introduction --
Contents note continued: Probability of Superiority: Independent Groups -- Introduction to Overlap and Related Measures -- Dominance Measure -- Cohen's Measures of Nonoverlap -- Relationships Among Measures of Effect Size -- Estimating Effect Sizes Throughout a Distribution -- Other Graphical Estimators of Effect Sizes -- Dependent Groups -- Summary -- Questions -- ch. 6 Effect Sizes for One-Way ANOVA and Nonparametric Approaches -- Introduction -- Assumptions -- ANOVA Results for This Chapter -- Standardized-Difference Measure of Overall Effect Size -- Standardized Overall Effect Size Using All Means -- Strength of Association -- Evaluation of Criticisms of Strength of Association -- Standardized-Difference Effect Sizes for Contrasts -- Worked Examples -- Unstandardized Differences Between Means -- More on Standardized Differences Between Means -- Intransitivity and the PS Effect Size -- Within-Groups Designs -- Summary -- Questions -- ch. 7 Effect Sizes for Factorial Designs --
Contents note continued: Introduction -- Assumptions and Handling Violations -- Discretizing Continuous Independent Variables -- Factors: Targeted, Peripheral, Extrinsic, Intrinsic -- Strength of Association: Proportion of Variance Explained -- Designs and Results for This Chapter -- Manipulated Factors Only -- Manipulated Targeted Factor and Intrinsic Peripheral Factor -- Illustrative Worked Examples -- Comparisons of Levels of a Manipulated Factor at One Level of a Peripheral Factor -- Targeted Classificatory Factor and Extrinsic Peripheral Factor -- Classificatory Factors Only -- Statistical Inference and Further Reading -- Within-Groups Factorial Designs -- Additional Designs, Measures, and Discussion -- Summary -- Questions -- ch. 8 Effect Sizes for Categorical Variables -- Introduction -- Unreliability of Categorization -- Dichotomizing a Continuous Variable -- Chi-Square Test and Phi -- Confidence Intervals and Null--Counternull Intervals for phipop --
Contents note continued: Difference Between Two Proportions -- Relative Risk -- Propensity-Score Analysis -- Relative Risk Reduction -- Number Needed to Treat -- Relationships Among Measures -- Odds Ratio -- Tables Larger Than 2 X 2 -- Multiway Tables -- More on Testing and Effect Sizes for Related Proportions -- Further Discussions -- Summary -- Questions -- ch. 9 Effect Sizes for Ordinal Categorical Dependent Variables (Rating Scales) -- Introduction -- Point-Biserial r Applied to Ordinal Categorical Data -- Probability of Superiority Applied to Ordinal Data -- Dominance Measure -- Somers' D, the Risk Difference, and the NNT -- Worked Example of the Dominance Statistic and NNTest -- Generalized Odds Ratio -- Cumulative Odds Ratios -- Phi Coefficient -- Further Reading -- Summary -- Questions -- ch. 10 Effect Sizes for Multiple Regression/Correlation -- Introduction -- Multiple Coefficient of Determination -- Semipartial Correlation -- Partial Correlation --
Contents note continued: Statistical Control of Unwanted Effects -- Higher-Order Correlation Coefficients -- More Statistical Significance Testing and Confidence Intervals -- Sets of Included and Excluded X Variables -- Multiple Regression and ANOVA: Dummy Coding -- Worked Example of Multiple Regression/Correlation -- Nonlinear Regression -- Hierarchical Linear Modeling (Multilevel Modeling) -- Path Analysis and Structural Equation Modeling -- Effect Size for Ordinal Multiple Regression -- Additional Topics and Reading -- Summary -- Questions -- ch. 11 Effect Sizes for Analysis of Covariance -- Introduction -- ANCOVA in Nonexperimental Research -- Proportion of Variance Explained Overall -- Proportion of Variance Explained by a Contrast -- Standardized Difference Between Means -- Unstandardized Difference Between Means -- Worked Examples of Effect Sizes -- Summary -- Questions -- ch. 12 Effect Sizes for Multivariate Analysis of Variance -- Introduction -- Assumptions of MANOVA --
Contents note continued: Statistical Tests -- Effect Sizes for One-Way MANOVA -- MANCOVA -- Factorial Between-Groups MANOVA -- One-Way Within-Groups MANOVA -- Effect Sizes for Mixed MANOVA Designs -- Effect Sizes for Within-Groups MANOVA Factorial Designs -- Additional Analyses -- Summary -- Questions.
Abstract:
"Noted for its comprehensive coverage, this greatly expanded new edition now covers the use of univariate and multivariate effect sizes. A variety of measures and estimators are reviewed along with their application, interpretation, and limitations. Noted for its practical approach, the book features numerous examples using real data for a variety of variables and designs, to help readers apply the material to their own data. Tips on the use of SPSS, SAS, R, And S-Plus are provided for the more tedious calculations. The book's broad disciplinary appeal results from its inclusion of a variety of examples from psychology, medicine, education, and other social sciences. Special attention is paid to confidence intervals, the statistical assumptions of the methods, and robust estimators of effect sizes. The extensive reference section is appreciated by all. With more than 40% new material, highlights of the new editon include:Three new multivariate chapters covering effect sizes for analysis of covariance, multiple regression/correlation, and multivariate analysis of variance.More learning tools in each chapter including introductions, summaries, "Tips and Pitfalls" and more conceptual and computational questions.More coverage of univariate effect sizes, confidence intervals, and effect sizes for repeated measures to reflect their increased use in research.More software references for calculating effect sizes and their confidence intervals including SPSS, SAS, R, and S-Plus.The data used in the book is now provided on the web along with suggested calculations for computational practice.Effect Sizes for Research, 2nd Edition covers standardized and unstandardized differences between means, correlational measures, strength of association, and parametric and nonparametric measures for between- and within-groups data. The book clearly demonstrates how the choice of an appropriate measure depends on such factors as whether variables are categorical, ordinal, or continuous; satisfying assumptions; sampling; and the source of variability in the population. Background information on multivariate statistics is provided for those who need it. Intended as a resource for professionals, researchers, and advanced students in a variety of fields, this book is also an excellent supplement for advanced statistics courses in psychology, education, the social sciences, business, and medicine. A prerequisite of introductory statistics through factorial analysis of variance and chi-square is recommended"--
Genre:
Added Author: