Quantitative Social Science: An Introduction in tidyverse

by Kosuke Imai, Nora Webb Williams

Overview

A tidyverse edition of the acclaimed textbook on data analysis and statistics for the social sciences and allied fields

Quantitative analysis is an essential skill for social science research, yet students in the social sciences and related areas typically receive little training in it. Quantitative Social Science is a practical introduction to data analysis and statistics written especially for undergraduates and beginning graduate students in the social sciences and allied fields, including business, economics, education, political science, psychology, sociology, public policy, and data science. Proven in classrooms around the world, this one-of-a-kind textbook engages directly with empirical analysis, showing students how to analyze and interpret data using the tidyverse family of R packages. Data sets taken directly from leading quantitative social science research illustrate how to use data analysis to answer important questions about society and human behavior.

• Emphasizes hands-on learning, not paper-and-pencil statistics
• Includes data sets from actual research for students to test their skills on
• Covers data analysis concepts such as causality, measurement, and prediction, as well as probability and statistical tools
• Features a wealth of supplementary exercises, including additional data analysis exercises and programming exercises
• Offers a solid foundation for further study
• Comes with additional course materials online, including notes, sample code, exercises and problem sets with solutions, and lecture slides

Product Details

ISBN-13:	9780691222271
Publisher:	Princeton University Press
Publication date:	08/23/2022
Pages:	488
Product dimensions:	7.00(w) x 10.00(h) x (d)

About the Author

Kosuke Imai is Professor of Government and of Statistics at Harvard University. Nora Webb Williams is Assistant Professor of Political Science at the University of Illinois, Urbana-Champaign.

Table of Contents

List of Tables xiii

List of Figures xv

Preface xvii

Preface to the Original Book xix

1 Introduction 1

1.1 Overview of the Book 3

1.2 How to Use This Book 7

1.3 Introduction to R and the tidyverse 8

1.3.1 Arithmetic Operations: R as a Calculator 9

1.3.2 R Scripts 10

1.3.3 Loading Packages 11

1.3.4 Objects 13

1.3.5 Vectors 15

1.3.6 Functions 17

1.3.7 Data Files: Loading and Subsetting 20

1.3.8 Adding Variables 27

1.3.9 Data Frames: Summarizing 28

1.3.10 Saving Objects 30

1.3.11 Loading Data in Other Formats 31

1.3.12 Programming and Learning Tips 32

1.4 Summary 33

1.5 Exercises 34

1.5.1 Bias in Self-Reported Turnout 34

1.5.2 Understanding World Population Dynamics 35

2 Causality 38

2.1 Racial Discrimination in the Labor Market 38

2.2 Subsetting Data in R 45

2.2.1 Logical Values and Operators 46

2.2.2 Relational Operators 48

2.2.3 Subsetting 49

2.2.4 Simple Conditional Statements 53

2.2.5 Factor Variables 53

2.3 Causal Effects and the Counterfactual 56

2.4 Randomized Controlled Trials 58

2.4.1 The Role of Randomization 59

2.4.2 Social Pressure and Voter Turnout 60

2.5 Observational Studies 65

2.5.1 Minimum Wage and Unemployment 65

2.5.2 Confounding Bias 68

2.5.3 Before-and-After and Difference-in-Differences Designs 71

2.6 Descriptive Statistics for a Single Variable 75

2.6.1 Quantiles 75

2.6.2 Standard Deviation 78

2.7 Summary 81

2.8 Exercises 82

2.8.1 Efficacy of Small Class Size in Early Education 82

2.8.2 Changing Minds on Gay Marriage 84

2.8.3 Success of Leader Assassination as a Natural Experiment 85

3 Measurement 88

3.1 Measuring Civilian Victimization during Wartime 88

3.2 Handling Missing Data in R 93

3.3 Visualizing the Univariate Distribution 96

3.3.1 Bar Plot 97

3.3.2 Histogram 100

3.3.3 Box Plot 103

3.3.4 Printing and Saving Graphs 105

3.4 Survey Sampling 106

3.4.1 The Role of Randomization 107

3.4.2 Nonresponse and Other Sources of Bias 111

3.5 Measuring Political Polarization 114

3.6 Summarizing Bivariate Relationships 116

3.6.1 Scatter Plot 116

3.6.2 Correlation 120

3.7 Quantile-Quantile Plot 124

3.8 Clustering 128

3.8.1 Matrix in R 128

3.8.2 List in R 130

3.8.3 The k-Means Algorithm 131

3.9 Summary 136

3.10 Exercises 137

3.10.1 Changing Minds on Gay Marriage: Revisited 137

3.10.2 Political Efficacy in China and Mexico 139

3.10.3 Voting in the United Nations General Assembly 141

4 Prediction 144

4.1 Predicting Election Outcomes 144

4.1.1 Loops in R 145

4.1.2 General Conditional Statements in R 148

4.1.3 Poll Predictions 152

4.2 Linear Regression 162

4.2.1 Facial Appearance and Election Outcomes 162

4.2.2 Correlation and Scatter Plots 165

4.2.3 Least Squares 166

4.2.4 Regression towards the Mean 173

4.2.5 Merging Data Sets in R 174

4.2.6 Model Fit 181

4.3 Regression and Causation 188

4.4 Randomized Experiments 188

4.4.1 Regression with Multiple Predictors 191

4.4.2 Heterogeneous Treatment Effects 197

4.4.3 Regression Discontinuity Design 203

4.5 Summary 209

4.6 Exercises 209

4.6.1 Prediction Based on Betting Markets 209

4.6.2 Election and Conditional Cash Transfer Program in Mexico 211

4.6.3 Government Transfer and Poverty Reduction in Brazil 214

5 Discovery 216

5.1 Textual Data 216

5.1.1 The Disputed Authorship of The Federalist Papers 216

5.1.2 Document-Term Matrix 221

5.1.3 Topic Discovery 223

5.1.4 Authorship Prediction 232

5.1.5 Cross-Validation 235

5.2 Network Data 238

5.2.1 Marriage Network in Renaissance Florence 238

5.2.2 Undirected Graph and Centrality Measures 240

5.2.3 Twitter-Following Network 245

5.2.4 Directed Graph and Centrality 247

5.3 Spatial Data 255

5.3.1 The 1854 Cholera Outbreak in London 256

5.3.2 Spatial Data in R 258

5.3.3 US Presidential Elections 264

5.3.4 Expansion of Walmart 268

5.3.5 Animation in R 270

5.4 Summary 272

5.5 Exercises 273

5.5.1 Analyzing the Preambles of Constitutions 273

5.5.2 International Trade Network 275

5.5.3 Mapping US Presidential Election Results over Time 277

6 Probability 279

6.1 Probability 279

6.1.1 Frequentist versus Bayesian 279

6.1.2 Definition and Axioms 281

6.1.3 Permutations 284

6.1.4 Sampling with and without Replacement 287

6.1.5 Combinations 289

6.2 Conditional Probability 291

6.2.1 Conditional, Marginal, and Joint Probabilities 291

6.2.2 Independence 301

6.2.3 Bayes' Rule 307

6.2.4 Predicting Race Using Surname and Residence Location 309

6.3 Random Variables and Probability Distributions 321

6.3.1 Random Variables 321

6.3.2 Bernoulli and Uniform Distributions 321

6.3.3 Binomial Distribution 325

6.3.4 Normal Distribution 328

6.3.5 Expectation and Variance 335

6.3.6 Predicting Election Outcomes with Uncertainty 339

6.4 Large Sample Theorems 342

6.4.1 The Law of Large Numbers 342

6.4.2 The Central Limit Theorem 345

6.5 Summary 350

6.6 Exercises 350

6.6.1 The Mathematics of Enigma 350

6.6.2 A Probability Model for Betting Market Election Prediction 352

6.6.3 Election Fraud in Russia 354

7 Uncertainty 357

7.1 Estimation 357

7.1.1 Unbiasedness and Consistency 358

7.1.2 Standard Error 366

7.1.3 Confidence Interval 371

7.1.4 Margin of Error and Sample Size Calculation in Polls 378

7.1.5 Analysis of Randomized Controlled Trials 383

7.1.6 Analysis Based on Student's t-Distribution 386

7.2 Hypothesis Testing 390

7.2.1 Tea-Tasting Experiment 390

7.2.2 The General Framework 394

7.2.3 One-Sample Tests 397

7.2.4 Two-Sample Tests 404

7.2.5 Pitfalls of Hypothesis Testing 409

7.2.6 Power Analysis 411

7.3 Linear Regression Model with Uncertainty 418

7.3.1 Linear Regression as a Generative Model 418

7.3.2 Unbiasedness of Estimated Coefficients 423

7.3.3 Standard Errors of Estimated Coefficients 426

7.3.4 Inference about Coefficients 428

7.3.5 Inference about Predictions 432

7.4 Summary 439

7.5 Exercises 439

7.5.1 Sex Ratio and the Price of Agricultural Crops in China 439

7.5.2 Filedrawer and Publication Bias in Academic Research 441

7.5.3 Analysis of the 1933 German Election during the Weimar Republic 443

8 Next 446

General Index 449

R Index 455