Behavioral Research Data Analysis with R (eBook)

Jonathan Baron, Yuelin Li (Autoren)

eBook Download: PDF

2011 | 2012
XII, 245 Seiten
Springer New York (Verlag)
978-1-4614-1238-0 (ISBN)

Infos

This book is written for behavioral scientists who want to consider adding R to their existing set of statistical tools, or want to switch to R as their main computation tool. The authors aim primarily to help practitioners of behavioral research make the transition to R. The focus is to provide practical advice on some of the widely-used statistical methods in behavioral research, using a set of notes and annotated examples. The book will also help beginners learn more about statistics and behavioral research. These are statistical techniques used by psychologists who do research on human subjects, but of course they are also relevant to researchers in others fields that do similar kinds of research.

The authors emphasize practical data analytic skills so that they can be quickly incorporated into readers' own research.

Yuelin Li is a research psychologist and a behavioral statistician.  His appointment at Memorial Sloan-Kettering Cancer Center allows him to apply a range of statistical techniques in understanding complex human behaviors---social network influence of young adult smoking, genetic-environment interaction in cognitive impairment, health behavior change, psychosocial and quality of life outcomes in  cancer treatment, survivorship, and end of life care.

Jonathan Baron is Professor of Psychology at the University of Pennsylvania, where he teaches Judgments and Decision and does research people's judgments and decisions about public policies.  He has been fascinated by the promise of computers since about 1960 and has come of age with them and used them in his research.  In 2000, he began the Web site (http://finzi.psych.upenn.edu) and document that led to this book, which was then mostly about data layout, until Yuelin Li (who shared the same PhD advisor, David Krantz) volunteered to help with the more substantive parts.  Baron is founding and current editor of the journal Judgment and Decision Making.

Inhalt

Behavioral Research Data Analysis with R 3
Preface 5
Contents 9
Chapter 1 Introduction 13
1.1 An Example R Session 13
1.2 A Few Useful Concepts and Commands 15
1.2.1 Concepts 15
1.2.2 Commands 16
1.2.2.1 Working Directory 16
1.2.2.2 Getting Help 17
1.2.2.3 Installing Packages 18
1.2.2.4 Assignment, Logic, and Arithmetic 18
1.2.2.5 Loading and Saving 20
1.2.2.6 Dealing with Objects 21
1.3 Data Objects and Data Types 21
1.3.1 Vectors of Character Strings 22
1.3.2 Matrices, Lists, and Data Frames 24
1.3.2.1 Summaries and Calculations by Row, Column, or Group 26
1.4 Functions and Debugging 27
Chapter 2 Reading and Transforming Data Format 30
2.1 Reading and Transforming Data 30
2.1.1 Data Layout 30
2.1.2 A Simple Questionnaire Example 30
2.1.2.1 Extracting Subsets of Data 31
2.1.2.2 Finding Means (or Other Things) of Sets of Variables 32
2.1.2.3 One Row Per Observation 32
2.1.3 Other Ways to Read in Data 36
2.1.4 Other Ways to Transform Variables 37
2.1.4.1 Contrasts 37
2.1.4.2 Averaging Items in a Within-Subject Design 38
2.1.4.3 Selecting Cases or Variables 39
2.1.4.4 Recoding and Replacing Data 39
2.1.4.5 Replacing Characters with Numbers 41
2.1.5 Using R to Compute Course Grades 41
2.2 Reshape and Merge Data Frames 42
2.3 Data Management with a SQL Database 44
2.4 SQL Database Considerations 46
Chapter 3 Statistics for Comparing Means and Proportions 49
3.1 Comparing Means of Continuous Variables 49
3.2 More on Manual Checking of Data 52
3.3 Comparing Sample Proportions 53
3.4 Moderating Effect in loglin() 55
3.5 Assessing Change of Correlated Proportions 59
3.5.1 McNemar Test Across Two Samples 60
Chapter 4 R Graphics and Trellis Plots 65
4.1 Default Behavior of Basic Commands 65
4.2 Other Graphics 66
4.3 Saving Graphics 66
4.4 Multiple Figures on One Screen 67
4.5 Other Graphics Tricks 67
4.6 Examples of Simple Graphs in Publications 68
4.6.1 http://journal.sjdm.org/8827/jdm8827.pdf 70
4.6.2 http://journal.sjdm.org/8814/jdm8814.pdf 73
4.6.3 http://journal.sjdm.org/8801/jdm8801.pdf 74
4.6.4 http://journal.sjdm.org/8319/jdm8319.pdf 75
4.6.5 http://journal.sjdm.org/8221/jdm8221.pdf 76
4.6.6 http://journal.sjdm.org/8210/jdm8210.pdf 78
4.7 Shaded Areas Under a Curve 79
4.7.1 Vectors in polygon() 81
4.8 Lattice Graphics 82
4.8.0.1 Mathematics Achievement and Socioeconomic Status 82
Chapter 5 Analysis of Variance: Repeated-Measures 88
5.1 Example 1: Two Within-Subject Factors 88
5.1.1 Unbalanced Designs 92
5.2 Example 2: Maxwell and Delaney 94
5.3 Example 3: More Than Two Within-Subject Factors 97
5.4 Example 4: A Simpler Design with Only One Within-Subject Variable 98
5.5 Example 5: One Between, Two Within 98
5.6 Other Useful Functions for ANOVA 100
5.7 Graphics with Error Bars 102
5.8 Another Way to do Error Bars Using plotCI() 104
5.8.1 Use Error() for Repeated-Measure ANOVA 105
5.8.1.1 Basic ANOVA Table with aov() 106
5.8.1.2 Using Error() Within aov() 107
5.8.1.3 The Appropriate Error Terms 107
5.8.1.4 Sources of the Appropriate Error Terms 108
5.8.1.5 Verify the Calculations Manually 110
5.8.2 Sphericity 111
5.8.2.1 Why Is Sphericity Important? 111
5.9 How to Estimate the Greenhouse–Geisser Epsilon? 112
5.9.1 Huynh–Feldt Correction 1
Chapter 6 Linear and Logistic Regression 117
6.1 Linear Regression 117
6.2 An Application of Linear Regression on Diamond Pricing 118
6.2.1 Plotting Data Before Model Fitting 119
6.2.2 Checking Model Distributional Assumptions 122
6.2.3 Assessing Model Fit 123
6.3 Logistic Regression 126
6.4 Log–Linear Models 127
6.5 Regression in Vector–Matrix Notation 128
6.6 Caution on Model Overfit and Classification Errors 130
Chapter 7 Statistical Power and Sample Size Considerations 136
7.1 A Simple Example 136
7.2 Basic Concepts on Statistical Power Estimation 137
7.3 t-Test with Unequal Sample Sizes 138
7.4 Binomial Proportions 139
7.5 Power to Declare a Study Feasible 140
7.6 Repeated-Measures ANOVA 140
7.7 Cluster-Randomized Study Design 142
Chapter 8 Item Response Theory 145
8.1 Overview 145
8.2 Rasch Model for Dichotomous Item Responses 145
8.2.1 Fitting a rasch() Model 146
8.2.2 Graphing Item Characteristics and Item Information 149
8.2.3 Scoring New Item Response Data 151
8.2.4 Person Fit and Item Fit Statistics 151
8.3 Generalized Partial Credit Model for Polytomous ItemResponses 152
8.3.1 Neuroticism Data 153
8.3.2 Category Response Curves and Item InformationCurves 153
8.4 Bayesian Methods for Fitting IRT Models 155
8.4.1 GPCM 155
8.4.2 Explanatory IRT 158
Chapter 9 Imputation of Missing Data 166
9.1 Missing Data in Smoking Cessation Study 166
9.2 Multiple Imputation with aregImpute() 168
9.2.1 Imputed Data 170
9.2.2 Pooling Results Over Imputed Datasets 171
9.3 Multiple Imputation with the mi Package 173
9.4 Multiple Imputation with the Amelia and Zelig Packages 176
9.5 Further Reading 178
Chapter 10 Linear Mixed-Effects Models in Analyzing Repeated-Measures Data 181
10.1 The "Language-as-Fixed-Effect Fallacy' 181
10.2 Recall Scores Example: One Between and One Within Factor 184
10.2.1 Data Preparations 184
10.2.2 Data Visualizations 185
10.2.3 Initial Modeling 186
10.2.4 Model Interpretation 186
10.2.4.1 Fixed Effects 186
10.2.4.2 Random Effects 189
10.2.5 Alternative Models 190
10.2.6 Checking Model Fit Visually 193
10.2.7 Modeling Dependence 194
10.3 Generalized Least Squares Using gls() 199
10.4 Example on Random and Nested Effects 202
10.4.1 Treatment by Therapist Interaction 204
Chapter 11 Linear Mixed-Effects Models in Cluster-Randomized Studies 209
11.1 The Television, School, and Family Smoking Prevention and Cessation Project 209
11.2 Data Import and Preparations 210
11.2.1 Exploratory Analyses 211
11.3 Testing Intervention Efficacy with Linear Mixed-Effects Models 214
11.4 Model Equation 217
11.5 Multiple-Level Model Equations 219
11.6 Model Equation in Matrix Notations 220
11.7 Intraclass Correlation Coefficients 224
11.8 ICCs from a Mixed-Effects Model 225
11.9 Statistical Power Considerationsfor a Group-Randomized Design 227
11.9.1 Calculate Statistical Power by Simulation 227
Appendix A Data Management with a Database 232
A.1 Create Database and Database Tables 232
A.2 Enter Data 233
A.3 Using RODBC to Import Data from an ACCESS Database 235
A.3.1 Step 1: Adding an ODBC Data Source Name 236
A.3.2 Step 2: ODBC Data Source Name Points to the ACCESS File 236
A.3.3 Step 3: Run RODBC to Import Data 237
References 239
Index 244

Produkt-Details

• Erscheint lt. Verlag: 2.12.2011
• Reihe/Serie: Use R!
• Reihe/Serie: Use R!
• Zusatzinfo: XII, 245 p. 32 illus., 11 illus. in color.
• Verlagsort: New York
• Sprache: englisch
• Themenwelt: Geisteswissenschaften ► Psychologie
• Themenwelt: Mathematik / Informatik ► Mathematik ► Statistik
• Themenwelt: Mathematik / Informatik ► Mathematik ► Wahrscheinlichkeit / Kombinatorik
• Themenwelt: Sozialwissenschaften ► Politik / Verwaltung
• Themenwelt: Sozialwissenschaften ► Soziologie ► Empirische Sozialforschung
• Themenwelt: Technik
• Schlagworte: behavioral science • R applications
• ISBN-10: 1-4614-1238-2 / 1461412382
• ISBN-13: 978-1-4614-1238-0 / 9781461412380