Atnaujinkite slapukų nuostatas

Statistical Power Analysis for the Social and Behavioral Sciences: Basic and Advanced Techniques [Kietas viršelis]

3.00/5 (3 ratings by Goodreads)
  • Formatas: Hardback, 386 pages, aukštis x plotis: 254x178 mm, weight: 900 g, 11 Tables, black and white
  • Išleidimo metai: 20-Nov-2013
  • Leidėjas: Routledge
  • ISBN-10: 1848729804
  • ISBN-13: 9781848729803
Kitos knygos pagal šią temą:
Statistical Power Analysis for the Social and Behavioral Sciences: Basic and Advanced TechniquesDidesnis paveikslėlis
  • Formatas: Hardback, 386 pages, aukštis x plotis: 254x178 mm, weight: 900 g, 11 Tables, black and white
  • Išleidimo metai: 20-Nov-2013
  • Leidėjas: Routledge
  • ISBN-10: 1848729804
  • ISBN-13: 9781848729803
Kitos knygos pagal šią temą:
Pastovi nuoroda: https://www.kriso.lt/db/9781848729803.html
Raktažodžiai:
"This will be the first book to demonstrate the application of power analysis to the newer more advanced techniques such as hierarchical linear modeling, meta-analysis, and structural equation modelling that are increasingly popular in behavioral and social science research"--

This is the first book to demonstrate the application of power analysis to the newer more advanced statistical techniques that are increasingly used in the social and behavioral sciences. Both basic and advanced designs are covered. Readers are shown how to apply power analysis to techniques such as hierarchical linear modeling, meta-analysis, and structural equation modeling. Each chapter opens with a review of the statistical procedure and then proceeds to derive the power functions. This is followed by examples that demonstrate how to produce power tables and charts. The book clearly shows how to calculate power by providing open code for every design and procedure in R, SAS, and SPSS. Readers can verify the power computation using the computer programs on the book's website. There is a growing requirement to include power analysis to justify sample sizes in grant proposals. Most chapters are self-standing and can be read in any order without much disruption.This book will help readers do just that. Sample computer code in R, SPSS, and SAS at www.routledge.com/9781848729810 are written to tabulate power values and produce power curves that can be included in a grant proposal.

Organized according to various techniques, chapters 1 – 3 introduce the basics of statistical power and sample size issues including the historical origin, hypothesis testing, and the use of statistical power int tests and confidence intervals. Chapters 4 - 6 cover common statistical procedures -- analysis of variance, linear regression (both simple regression and multiple regression), correlation, analysis of covariance, and multivariate analysis. Chapters 7 - 11 review the new statistical procedures -- multi-level models, meta-analysis, structural equation models, and longitudinal studies. The appendixes contain a tutorial about R and show the statistical theory of power analysis.

Intended as a supplement for graduate courses on quantitative methods, multivariate statistics, hierarchical linear modeling (HLM) and/or multilevel modeling and SEM taught in psychology, education, human development, nursing, and social and life sciences, this is the first text on statistical power for advanced procedures. Researchers and practitioners in these fields also appreciate the book‘s unique coverage of the use of statistical power analysis to determine sample size in planning a study. A prerequisite of basic through multivariate statistics is assumed.

Recenzijos

"This book extends earlier landmark texts by adding sample-size estimation for multilevel and longitudinal designs, meta-analysis, and structural-equation modeling. It is written thoughtfully and understandably. Readers will benefit enormously from the inclusion of computer code (in R, SAS and SPSS) for conducting the power analyses described. I recommend the book very highly to any researcher who wants to design research in the social sciences." - John B. Willett, Harvard University, USA

"The author skillfully blends simple explanations of core concepts with more advanced material in a way that will make the work attractive to a range of readers in psychology and related disciplines. This text will be useful for postgraduate quantitative methods courses and for researchers. The coverage - from t tests through to multilevel models and SEM - is impressive. I found the examples of R, SPSS, and SAS code invaluable." - Thom Baguley, Nottingham Trent University, UK "This is a long-awaited, comprehensive book on power analysis after Cohens (1988) seminal book. The updated content accompanied by sample computer code is well suited for quantitative researchers in the social and behavioral sciences." - Wei Pan, Duke University, USA

"This book provides a more comprehensive treatment of power analysis than any other work. ... This is likely to be the "go to" book for more complex designs. ... I found the writing style clear. ... The primary audience for this book would be all investigators who seek external funding for their work." - Warren W. Tryon, Fordham University, USA

"This book would be good for departments of psychology, sociology, social work, nursing and public health. Most of the PhD programs in these departments have an advanced research methods course that could use this book. ... The Cohen book has been the standard in the field for over 20 years. ... This book would make a very nice update on a classic." - Jay Maddock, University of Hawaii, USA

"This book is] a valuable extension beyond what is currently provided by other books on power. This book would contribute significantly to the field, most notably by covering the advanced and more complex techniques. Liu and his work are well known in this field.[ This] bookcould serve as the primary text for a course on power. ...This book would basically have the field to itself." Geoff Cumming, La Trobe University, Australia

List of Figures
ix
List of Tables
xi
Preface xiii
1 Introduction
1(16)
1.1 Scientific Research
2(2)
1.2 Origin of Statistical Power
4(1)
1.3 Some Preliminaries
5(12)
1.3.1 Distribution
6(7)
1.3.2 Expectation
13(2)
1.3.3 Variance
15(2)
2 Statistical Power
17(38)
2.1 Hypothesis Testing
17(4)
2.2 Statistical Power
21(11)
2.2.1 Statistical Power of a z-test
21(6)
2.2.2 Effect Size
27(3)
2.2.3 Sample Size
30(2)
2.3 Statistical Power and Sample Size for t Tests
32(23)
2.3.1 One Sample t Test
33(3)
2.3.2 Two Sample Independent t Test
36(7)
2.3.3 Dependent t Test
43(7)
2.3.4 Uncertainty in Power Analysis
50(5)
3 Power of Confidence Interval
55(20)
3.1 Sample Size and Confidence Interval in a z-test
56(1)
3.2 Sample Size and Confidence Interval in a t Test
57(4)
3.3 Statistical Power and Confidence Interval
61(10)
3.3.1 Statistical Power and Confidence Interval Width in a z-test
62(2)
3.3.2 Statistical Power and Confidence Interval Width in a Test
64(7)
3.4 Derivation of Π1 and ω
71(4)
4 Analysis of Variance
75(36)
4.1 One-way Analysis of Variance
75(17)
4.1.1 Non-centrality Parameter in ANOVA
77(4)
4.1.2 Statistical Power for the F Test
81(4)
4.1.3 Statistical Power for Contrast Tests
85(7)
4.2 Two-way Analysis of Variance
92(12)
4.2.1 Statistical Power for Testing Main Effects
92(5)
4.2.2 Statistical Power for Testing Interaction
97(7)
4.3 Random-effects ANOVA
104(7)
5 Linear Regression
111(26)
5.1 Simple Regression
112(7)
5.2 Testing Correlation r
119(4)
5.3 Multiple Regression
123(6)
5.4 Analysis of Covariance
129(8)
5.4.1 Statistical Power with Covariate Adjustment
129(4)
5.4.2 Small Sample Size and Standard Error with Covariate Adjustment
133(4)
6 Multivariate Analysis
137(32)
6.1 Hotelling T2 for Two-group Comparison
137(8)
6.2 Multivariate Analysis of Variance
145(14)
6.3 Multivariate Analysis of Covariance
159(10)
7 Multi-level Models
169(46)
7.1 Cluster Randomized Trials
170(21)
7.1.1 Two Treatment Arms
170(8)
7.1.2 Three Treatment Arms
178(13)
7.2 Multi-site Randomized Trials
191(24)
7.2.1 Two Treatment Conditions at Each Site
191(9)
7.2.2 Multiple Treatment Conditions at Each Site
200(12)
7.2.3 Derivation of the F Statistic
212(3)
8 Complex Multi-level Models
215(38)
8.1 Three-level Cluster Randomized Trial
215(9)
8.2 Multi-site Cluster Randomized Trial
224(11)
8.3 Covariate Adjustment
235(8)
8.4 Optimal Sample Allocation
243(10)
8.4.1 Cluster Randomized Trial
244(3)
8.4.2 Multi-site Randomized Trial
247(2)
8.4.3 Unequal Costs Per Unit of Randomization
249(4)
9 Meta-analysis
253(16)
9.1 Fixed-Effects Meta-analysis
254(6)
9.1.1 Test for Main Effect
254(5)
9.1.2 Test for Homogeneity of Individual Effect Sizes ...
259(1)
9.2 Random-Effects Meta-analysis
260(9)
9.2.1 Test for Main Effect
260(6)
9.2.2 Test for the Variance of Individual Effect Sizes
266(3)
10 Structural Equation Models
269(30)
10.1 Path Analysis
270(16)
10.1.1 Statistical Power for Testing Individual Equations and Direct Effects
272(1)
10.1.2 Statistical Power in Path Analysis
273(13)
10.2 Factor Analysis
286(3)
10.3 Structural Equation Model
289(10)
10.3.1 Statistical Power in SEM
291(8)
11 Longitudinal Studies
299(48)
11.1 Multivariate Analysis of Repeated Measures
300(15)
11.1.1 Linear Change
300(8)
11.1.2 Quadratic Change
308(7)
11.2 HLM: Fixed Coefficient for Slope
315(9)
11.3 HLM: Random Coefficients Model
324(23)
11.3.1 Linear Change
325(6)
11.3.2 Quadratic Change
331(8)
11.3.3 Orthogonal Polynomial Model
339(4)
11.3.4 Cubic Change
343(4)
Appendix A Cumulative Distribution Function
347(4)
A.1 Cumulative Distribution Function and Quantile
347(1)
A.2 Non-central Cumulative Distribution Function
348(3)
Appendix B R Tutorial
351(14)
B.1 Arithmetic
351(2)
B.2 Data
353(1)
B.3 Loop
354(1)
B.4 Statistics
355(2)
B.5 Simulation
357(2)
B.6 Graphics
359(1)
B.7 Function
360(1)
B.8 Statistical Analysis
361(4)
B.8.1 Basic Statistics
361(1)
B.8.2 Helmert Contrast
362(3)
Bibliography 365(8)
Index 373
Xiaofeng Steven Liu is an Associate Professor at the University of South Carolina.