Research design and analysis : a primer for the non-statistician / Leslie D. Rosenstein.

ABOUT THE AUTHOR
LESLIE D. ROSENSTEIN, PHD, serves on the faculty of the Department of Psychiatry, Division of Psychology at the University of Texas Southwestern Medical Center. She earned her PhD in??Differential Psychology from the University of Texas at Austin and completed a doctoral respecialization in Clinical Neuropsychology at the University of Arizona in Tucson. Dr. Rosenstein has taught courses in??Introductory Statistics and Research Design and Analysis to Psychology students.

Includes bibliographical references and index.

TABLE OF CONTENTS
List of Figures xiii

List of Tables xv

Introduction xix

Section 1 The Purpose, Ethics, and Rules of Research 1

1 The Purpose and Ethics of Research 3

1.1 The Purpose and Risks of Research 3

1.2 History of Harm to Humans 4

1.3 Ethical Issues in the Social Sciences 9

1.4 History of Harm to Animal Subjects in Research 10

1.4.1 Summary 12

1.5 Ethics, Principles, and Guidelines 12

1.6 Statutes and Regulations Protecting Humans and Animals in Research 16

1.7 More About Informed Consent 18

1.8 The Importance of Freedom to Withdraw 22

1.9 Separation of Provider–Researcher Role 22

1.10 Undue Influence 24

1.11 Anonymity 24

1.12 Summary 25

Section 2 Basic Research Designs and Validity 27

2 Research Validity 29

2.1 Internal Validity 30

2.1.1 History 30

2.1.2 Maturation 31

2.1.3 Measurement Error 32

2.1.4 Selection Bias and Random Assignment 33

2.1.5 Attrition 35

2.1.6 Experimenter Bias 35

2.1.7 Expectation 36

2.1.8 Sensitization and Practice Effects 36

2.1.9 Incorrect Conclusions of Causality 37

2.2 External Validity 37

2.3 Summary 45

3 Research Designs 47

3.1 The Lingo 47

3.2 Between‐Subjects Designs 49

3.2.1 More Examples of Between‐Subjects Designs 49

3.2.2 Statistical Analyses for Between‐Subjects Designs 50

3.3 Within‐Subjects Designs/Repeated Measures 52

3.3.1 Statistical Analyses for Within‐Subjects Designs 53

3.4 Between–Within Subjects Designs (Mixed Factorial/Split‐Plot Designs) 54

3.4.1 Statistical Analyses for Between–Within Subjects Designs 55

3.5 Latin Square Designs 57

3.5.1 Summary 59

3.5.2 Double Latin Square Designs 59

3.5.3 Graeco‐Latin and Hyper Graeco‐Latin Square Designs 59

3.6 Nesting 60

3.7 Matching 60

3.8 Blocking 61

3.9 Nonexperimental Research 62

3.10 Case Studies 62

3.11 Summary 64

Section 3 The Nuts and Bolts of Data Analysis 65

4 Interpretation 67

4.1 Probability and Significance 67

4.2 The Null Hypothesis, Type I (α), and Type II (β) Errors 68

4.3 Power 69

4.4 Managing Error Variance to Improve Power 71

4.5 Power Analyses 72

4.6 Effect Size 72

4.7 Confidence Intervals and Precision 74

4.8 Summary 76

5 Parametric Statistical Techniques 77

5.1 A Little More Lingo 77

5.1.1 Population Parameters Versus Sample Statistics 78

5.1.2 Data 78

5.1.2.1 Ratio and Interval Data 78

5.1.2.2 Ordinal Data 78

5.1.2.3 Nominal Data 79

5.1.3 Central Tendency 79

5.1.3.1 Mode 79

5.1.3.2 Median 79

5.1.3.3 Mean 86

5.1.4 Distributions 86

5.1.5 Dependent Variables 92

5.1.5.1 To Scale or Not to Scale 95

5.1.6 Summary 97

5.2 t Tests 97

5.2.1 Independent Samples t Tests 97

5.2.2 Matched Group Comparison 98

5.2.3 Assumptions of t Tests 99

5.2.4 More Examples of Studies Employing t Tests 100

5.2.5 Statistical Software Packages for Conducting t Tests 101

5.3 The NOVAs and Mixed Linear Model Analysis 101

5.3.1 ANOVA 102

5.3.1.1 ANOVA with a Multifactorial Design 104

5.3.1.2 Main Effects and Interactions 104

5.3.1.3 More Illustrations of Interactions and Main Effects 106

5.3.1.4 Assumptions of ANOVA 107

5.3.2 ANCOVA 109

5.3.3 MANOVA/MANCOVA 111

5.3.4 Statistical Software Packages for Conducting ANOVA/ANCOVA/MANOVA 114

5.3.5 Repeated Measures: ANOVA‐RM and Mixed Linear Model Analysis 114

5.3.5.1 ANOVA‐RM 114

5.3.5.2 Mixed Linear Model Analysis 116

5.3.5.3 ANCOVA 117

5.3.5.4 Statistical Software Packages for Conducting Repeated Measures Analyses 117

5.3.6 Summary 119

5.4 Correlation and Regression 120

5.4.1 Correlation and Multiple Correlation 120

5.4.2 Regression and Multiple Regression 121

5.4.3 Statistical Software Packages for Conducting Correlation and Regression 124

5.5 Logistic Regression 126

5.5.1 Statistical Software Packages for Conducting Logistic Regression 128

5.6 Discriminant Function Analysis 128

5.6.1 Statistical Software Packages for Conducting Discriminant Function Analysis 128

5.7 Multiple Comparisons 129

5.8 Summary 131

6 Nonparametric Statistical Techniques 133

6.1 Chi‐Square 134

6.1.1 Statistical Software Packages for Conducting Chi‐Square 136

6.2 Median Test 137

6.2.1 Statistical Software Packages for Conducting Median Tests 137

6.3 Phi Coefficient 137

6.3.1 Statistical Software Packages for Calculating the Phi Coefficient 139

6.4 Mann–Whitney U Test (Wilcoxon Rank Sum Test) 139

6.4.1 Statistical Software Packages for Conducting a Mann–Whitney U Test 141

6.5 Sign Test and Wilcoxon Signed‐rank Test 142

6.5.1 Statistical Software Packages for Conducting Sign Tests 143

6.6 Kruskal–Wallis Test 144

6.6.1 Statistical Software Packages for Conducting a Kruskal–Wallis Test 144

6.7 Rank‐Order Correlation 145

6.7.1 Statistical Software Packages for Conducting Rank‐order Correlations 146

6.8 Summary 147

7 Meta‐Analytic Studies 149

7.1 The File Drawer Effect 150

7.2 Analyzing the Meta‐Analytic Data 151

7.3 How to Read and Interpret a Paper Reporting a Meta‐Analysis 153

7.4 Statistical Software Packages for Conducting Meta‐Analyses 155

7.5 Summary 155

Section 4 Reporting, Understanding, and Communicating Research Findings 157

8 Disseminating Your Research Findings 159

8.1 Preparing a Research Report 159

8.2 Presenting Your Findings at a Conference 167

8.3 Summary 168

9 Concluding Remarks 169

9.1 Why is it Important to Understand Research Design and Analysis as a Consumer? 169

9.2 Research Ethics and Responsibilities of Journalists 175

9.3 Responsibilities of Researchers 177

9.4 Conclusion 178

Appendix A Data Sets and Databases 179

Appendix B Statistical Analysis Packages 195

Appendix C Helpful Statistics Resources 217

Glossary 221

References 233

Index 243

"This book helps readers understand research design and analysis. The intended audience includes those in the sciences who wish to conduct their own research without investing several semesters completing coursework in statistics and related fields, as well as those in the sciences and media who wish to read published research in an informed manner. In the former case, this book will provide a general basis for designing and conducting research, though with the assistance of a statistical consultant. The chapters of this book are laid out into three major sections. In the first section, the author briefly reviews the purpose of research as well as ethics and rules guiding research involving human participants and animal subjects. In the second section, the author walks the reader through basic research designs and validity. In the third section, the author provides a cursory review of statistics, including interpretation and specific techniques. This includes a description of when each statistical technique is appropriate and informative. The section on statistical techniques is intended to help the new researcher to be able to communicate in an informed manner with a statistical consultant or to be able to comprehend the jargon they may find in many research documents. Finally, the author reviews the how-to's of reporting and presenting research findings"-- Provided by publisher.

Description based on print version record and CIP data provided by publisher.