Atnaujinkite slapukų nuostatas

Applied Power Analysis for the Behavioral Sciences: 2nd Edition 2nd edition [Minkštas viršelis]

  • Formatas: Paperback / softback, 194 pages, aukštis x plotis: 229x152 mm, weight: 300 g, 93 Tables, black and white
  • Išleidimo metai: 01-Feb-2019
  • Leidėjas: Routledge
  • ISBN-10: 1138044598
  • ISBN-13: 9781138044593
Kitos knygos pagal šią temą:
Applied Power Analysis for the Behavioral Sciences: 2nd Edition 2nd editionDidesnis paveikslėlis
  • Formatas: Paperback / softback, 194 pages, aukštis x plotis: 229x152 mm, weight: 300 g, 93 Tables, black and white
  • Išleidimo metai: 01-Feb-2019
  • Leidėjas: Routledge
  • ISBN-10: 1138044598
  • ISBN-13: 9781138044593
Kitos knygos pagal šią temą:
Pastovi nuoroda: https://www.kriso.lt/db/9781138044593.html
Raktažodžiai:

Applied Power Analysis for the Behavioral Sciences is a practical "how-to" guide to conducting statistical power analyses for psychology and related fields. The book provides a guide to conducting analyses that is appropriate for researchers and students, including those with limited quantitative backgrounds. With practical use in mind, the text provides detailed coverage of topics such as how to estimate expected effect sizes and power analyses for complex designs. The topical coverage of the text, an applied approach, in-depth coverage of popular statistical procedures, and a focus on conducting analyses using R make the text a unique contribution to the power literature. To facilitate application and usability, the text includes ready-to-use R code developed for the text. An accompanying R package called pwr2ppl (available at https://github.com/chrisaberson/pwr2ppl) provides tools for conducting power analyses across each topic covered in the text.

List of Figures xi
List of Tables xii
Preface xvi
Overview of the Book
xvii
What is New in this Edition?
xvii
Formulae and Calculations
xvii
Approaches to Power
xviii
The pwr2ppl Companion Package
xviii
Acknowledgments xx
1 What is Power? Why is Power Important? 1(17)
Review of Null Hypothesis Significance Testing
1(1)
Effect Sizes and Their Interpretation
2(1)
What Influences Power?
3(3)
Central and Noncentral Distributions
6(2)
Misconceptions about Power
8(1)
Empirical Reviews of Power
9(1)
Consequences of Underpowered Studies
10(1)
Overview of Approaches to Determining Effect Size for Power Analysis
11(3)
Post Hoc Power (a.k.a. Observed or Achieved Power)
14(2)
How Much Power?
16(1)
Summary
17(1)
2 Chi Square and Tests for Proportions 18(16)
Necessary Information
18(1)
Factors Affecting Power
19(1)
Key Statistics
19(1)
Example 2.1: 2 x 2 Test of Independence
20(6)
Example 2.2: 2 x 2 Chi Square Test for Independence Using R
26(1)
Example 2.3: Other x 2 Tests
27(1)
Example 2.4: General Effect Size-Based Approaches Using R
27(1)
Tests for Single Samples and Independent Proportions
28(1)
Example 2.5: Single Sample Comparison
29(2)
Example 2.6: Independent Proportions Comparison
31(1)
Additional Issues
32(1)
Summary
33(1)
3 Independent Samples and Paired t-tests 34(20)
Necessary Information
34(1)
Factors Affecting Power
35(1)
Key Statistics
36(2)
A Note about Effect Size for Two-Group Comparisons
38(1)
Example 3.1: Comparing Two Independent Groups
39(3)
Example 3.2: Power for Independent Samples t using R
42(1)
Example 3.3: Paired t-test
43(1)
Example 3.4: Power for Paired t using R
44(1)
Example 3.5: Power from Effect Size Estimate
44(1)
Dealing with Unequal Variances, Unequal Sample Sizes, and Violation of Assumptions
45(4)
Example 3.6: Unequal Variances and Unequal Sample Sizes
49(3)
Additional Issues
52(1)
Summary
53(1)
4 Correlations and Differences between Correlations 54(15)
Necessary Information
54(1)
Factors Affecting Power
54(1)
Zero-Order Correlation
55(1)
Example 4.1: Zero-order Correlations
55(2)
Comparing Two Independent Correlations
57(1)
Example 4.2: Comparing Independent Correlations
58(2)
Comparing Two Dependent Correlations (One Variable in Common)
60(1)
Example 4.3: Comparing Dependent Correlations, One Variable in Common
61(2)
Comparing Two Dependent Correlations (No Variables in Common)
63(1)
Example 4.4: Comparing Dependent Correlations, No Variables in Common
64(3)
Note on Effect Sizes for Comparing Correlations
67(1)
Additional Issues
67(1)
Summary
68(1)
5 Between Subjects ANOVA (One and Two Factors) 69(19)
Necessary Information
69(1)
Factors Affecting Power
69(1)
Omnibus Versus Contrast Power
70(1)
Key Statistics
70(2)
Example 5.1: One Factor ANOVA
72(2)
Example 5.2: One Factor ANOVA with Orthogonal Contrasts
74(4)
ANOVA with Two Factors
78(1)
Example 5.3: Two Factor ANOVA with Interactions
79(4)
Power for Multiple Effects
83(3)
Additional Issues
86(1)
Summary
87(1)
6 Within Subjects Designs with ANOVA and Linear Mixed Models 88(12)
Necessary Information
88(1)
Factors Affecting Power
88(2)
Key Statistics
90(1)
Example 6.1: One Factor Within Subjects Design
91(2)
Example 6.2: Sphericity Adjustments
93(1)
Example 6.3: Linear Mixed Model Approach to Repeated Measures
93(1)
Example 6.4: A Serious Sphericity Problem
94(1)
Trend Analysis
94(1)
Example 6.5: Trend Analysis
95(1)
Example 6.6: Two Within Subject Factors Using ANOVA
96(1)
Example 6.7: Simple Effects Using ANOVA
97(1)
Example 6.8: Two Factor Within and Simple Effects Using LMM
98(1)
Additional Issues
99(1)
Summary
99(1)
7 Mixed Model ANOVA and Multivariate ANOVA 100(12)
Necessary Information
100(1)
Factors Affecting Power
100(1)
Key Statistics
101(1)
ANOVA with Between and Within Subject Factors
101(1)
Example 7.1: ANOVA with One Within Subjects Factor and One Between Subjects Factor
101(2)
Example 7.2: Linear Mixed Model with One Within Subjects Factor and One Between Subjects Factor
103(1)
Multivariate ANOVA
104(3)
Example 7.3: Multivariate ANOVA
107(2)
Additional Issues
109(2)
Summary
111(1)
8 Multiple Regression 112(23)
Necessary Information
112(1)
Factors Affecting Power
112(2)
Key Statistics
114(3)
Example 8.1: Power for a Two Predictor Model (R2 Model and Coefficients)
117(4)
Example 8.2: Power for Three Predictor Models
121(1)
Example 8.3: Power for Detecting Differences between Two Dependent Coefficients
122(3)
Example 8.4: Power for Detecting Differences between Two Independent Coefficients
125(2)
Example 8.5: Comparing Two Independent R2 Values
127(1)
Multiplicity and Direction of Predictor Correlations
128(4)
Example 8.6: Power(All) with Three Predictors
132(1)
Additional Issues
133(1)
Summary
134(1)
9 Analysis of Covariance, Moderated Regression, Logistic Regression, and Mediation 135(22)
Analysis of Covariance
135(1)
Example 9.1: ANCOVA
136(3)
Moderated Regression Analysis (Regression with Interactions)
139(4)
Example 9.2: Regression Analogy (Coefficients)
143(1)
Example 9.3: Regression Analogy (R2 Change)
144(1)
Example 9.4: Comparison on Correlations/Simple Slopes
145(2)
Logistic Regression
147(1)
Example 9.5: Logistic Regression with a Single Categorical Predictor
148(1)
Example 9.6: Logistic Regression with a Single Continuous Predictor
149(2)
Example 9.7: Power for One Predictor in a Design with Multiple Predictors
151(1)
Mediation (Indirect Effects)
152(1)
Example 9.8: One Mediating Variable
153(1)
Example 9.9: Multiple Mediating Variables
154(1)
Additional Issues
155(1)
Summary
156(1)
10 Precision Analysis for Confidence Intervals 157(16)
Necessary Information
158(1)
Confidence Intervals
158(1)
Types of Confidence Intervals
159(1)
Example 10.1: Confidence Limits around Differences between Means
159(2)
Determining Levels of Precision
161(2)
Confidence Intervals around Effect Sizes
163(1)
Example 10.2: Confidence Limits around d
163(2)
Precision for a Correlation
165(1)
Example 10.3: Confidence Limits around r
165(2)
Example 10.4: Precision for R2
167(2)
Supporting Null Hypotheses
169(1)
Example 10.5: "Supporting" Null Hypotheses
169(1)
Additional Issues
170(1)
Summary
171(2)
11 Additional Issues and Resources 173(8)
Accessing the Analysis Code
173(1)
Using Loops to Get Power for a Range of Values
173(1)
How to Report Power Analyses
174(1)
Example 11.1: Reporting a Power Analysis for a Chi-Square Analysis
175(1)
Example 11.2: Reporting a Power Analysis for Repeated Measures ANOVA
175(1)
Reporting Power if Not Addressed A Priori
175(1)
Statistical Test Assumptions
176(1)
Effect Size Conversion Formulae
176(1)
General (Free) Resources for Power and Related Topics
177(1)
Resources for Additional Analyses
178(1)
Improving Power without Increasing Sample Size or Cost
179(2)
References 181(7)
Index 188
Chris Aberson is Professor of Psychology at Humboldt State University. His research interests in in social psychology include prejudice, racism, and attitudes toward affirmative action. His quantitative interests focus on statistical power. He serves as Editor-in-Chief for Analyses of Social Issues and Public Policy.