Methods Matter Improving Causal Inference in Educational and Social Science Research
Buku ini diterbitkan tahun 2011 Oleh Oxford University Press., adalah buku edisi Pertama.
Judul: Methods Matter Improving Causal Inference
in Educational and Social Science Research
Oleh: Richard J. Murnane, et al
Penerbit: Oxford University Press.
Tahun: 2011
Jumlah Halaman: 414 hal.
Penulis:
Richard J. Murnane
John B. Willett
Lingkup Pembahasan:
Buku ini mengemukakan 14 pokok bahasan utama, yaitu 1 Tantangan Penelitian Pendidikan,
2 Pentingnya Teori, 3 Merancang Penelitian untuk Menetapkan Pertanyaan kausal, 4 Penyidik-Merancang Percobaan Acak, 5 Tantangan dalam Merancang, Pelaksana, dan Belajar dari Percobaan Acak, 6 statistik Power dan Sampel, 7 Experimental Research Ketika Peserta Apakah Clustered Dalam Grup Utuh, 8 Menggunakan Percobaan Alam Memberikan Penyelesaian Variabilitas "Diperdebatkan eksogen", 9 Memperkirakan kausal Efek Menggunakan Pendekatan Regresi-Diskontinuitas, 10 Memperkenalkan Instrumental-Variabel Estimasi, 11 Menggunakan IVE untuk Recover Pengaruh Perlakuan dalam Kuasi-Eksperimen, 12 Berurusan dengan Bias di Treatment Efek Perkiraan dari data Non eksperimental, 13 Pelajaran metodologis dari Long Quest, dan 14 Pelajaran substantif dan Pertanyaan Baru.
Daftar Isi:
Preface xi
1 The Challenge for Educational Research 3
The Long Quest 3
The Quest Is Worldwide 9
What This Book Is About 10
What to Read Next 12
2 The Importance of Theory 14
What Is Theory? 15
Theory in Education 19
Voucher Theory 21
What Kind of Theories? 24
What to Read Next 24
3 Designing Research to Address Causal Questions 26
Conditions to Strive for in All Research 27
Making Causal Inferences 29
Past Approaches to Answering Causal Questions in Education 31
The Key Challenge of Causal Research 33
What to Read Next 39
4 Investigator-Designed Randomized Experiments 40
Conducting Randomized Experiments 41
The Potential Outcomes Framework 41
An Example of a Two-Group Experiment 45
Analyzing Data from Randomized Experiments 48
The Better Your Research Design, the Simpler Your Data Analysis 48
Bias and Precision in the Estimation of Experimental Effects 52
What to Read Next 60
5 Challenges in Designing, Implementing, and Learning from Randomized Experiments 61
Critical Decisions in the Design of Experiments 62
Defi ning the Treatment 64
Defi ning the Population from Which Participants Will Be Sampled 66
Deciding Which Outcomes to Measure 67
Deciding How Long to Track Participants 68
Threats to the Validity of Randomized Experiments 69
Contamination of the Treatment–Control Contrast 70
Cross-overs 70
Attrition from the Sample 71
Participation in an Experiment Itself Affects Participants’ Behavior 73
Gaining Support for Conducting Randomized Experiments: Examples from India 74
Evaluating an Innovative Input Approach 75
Evaluating an Innovative Incentive Policy 79
What to Read Next 81
6 Statistical Power and Sample Size 82
Statistical Power 83
Reviewing the Process of Statistical Inference 83
Defi ning Statistical Power 92
Factors Affecting Statistical Power 96
The Strengths and Limitations of Parametric Tests 101
The Benefi ts of Covariates 102
The Reliability of the Outcome Measure Matters 103
The Choice Between One-Tailed and Two-Tailed Tests 105
What to Read Next 106
7 Experimental Research When Participants Are Clustered Within Intact Groups 107
Random-Intercepts Multilevel Model to Estimate Effect Size When Intact Groups Are
Randomized to Experimental Conditions 110
Statistical Power When Intact Groups of Participants Are Randomized to Experimental Conditions 120
Statistical Power of the Cluster-Randomized Design and Intraclass Correlation 122
Fixed-Effects Multilevel Models to Estimate Effect Size When Intact Groups of
Participants Are Randomized to Experimental Conditions 128
Specifying a Fixed-Effects Multilevel Model 128
Choosing Between Random- and Fixed-Effects Specifi cations 131
What to Read Next 134
8 Using Natural Experiments to Provide “Arguably Exogenous” Treatment Variability 135
Natural- and Investigator-Designed Experiments: Similarities and Differences 136
Two Examples of Natural Experiments 137
The Vietnam-Era Draft Lottery 137
The Impact of an Offer of Financial Aid for College 141
Sources of Natural Experiments 145
Choosing the Width of the Analytic Window 150
Threats to Validity in Natural Experiments with a Discontinuity Design 152
Accounting for the Relationship Between the Outcome and the Forcing Variable in a
Discontinuity Design 153
Actions by Participants Can Undermine Exogenous Assignment to Experimental
Conditions in a Natural Experiment with a Discontinuity Design 163
What to Read Next 164
9 Estimating Causal Effects Using a Regression-Discontinuity Approach 165
Maimonides’ Rule and the Impact of Class Size on Student Achievement 166
A Simple First-Difference Analysis 170
A Difference-in-Differences Analysis 171
A Basic Regression-Discontinuity Analysis 174
Choosing an Appropriate Bandwidth 181
Generalizing the Relationship Between the Outcome and the Forcing Variable 186
Specifi cation Checks Using Pseudo-Outcomes and Pseudo–Cut-offs 192
Regression-Discontinuity Designs and Statistical Power 195
Additional Threats to Validity in a Regression-Discontinuity Design 197
What to Read Next 202
10 Introducing Instrumental-Variables Estimation 203
Introducing Instrumental-Variables Estimation 204
Bias in the OLS Estimate of the Causal Effect of Education on Civic Engagement 206
Instrumental-Variables Estimation 215
Two Critical Assumptions That Underpin Instrumental-Variables Estimation 223
Alternative Ways of Obtaining the Instrumental-Variables Estimate 226
Obtaining an Instrumental-Variables Estimate by the Two-Stage Least-Squares Method 227
Obtaining an Instrumental-Variables Estimate by Simultaneous-Equations Estimation 233
Extensions of the Basic Instrumental-Variable Estimation Approach 238
Incorporating Exogenous Covariates into Instrumental-Variable Estimation 238
Incorporating Multiple Instruments into the First-Stage Model 243
Examining the Impact of Interactions Between the Endogenous Question
Predictor and Exogenous Covariates in the Second-Stage Model 247
Choosing Appropriate Functional Forms for Outcome/Predictor
Relationships in First- and Second-Stage Models 251
Finding and Defending Instruments 252
Proximity of Educational Institutions 253
Institutional Rules and Personal Characteristics 257
Deviations from Cohort Trends 261
The Search Continues 263
What to Read Next 264
11 Using IVE to Recover the Treatment Effect in a Quasi-Experiment 265
The Notion of a “Quasi-Experiment” 267
Using IVE to Estimate the Causal Impact of a Treatment in a Quasi-Experiment 269
Further Insight into the IVE (LATE) Estimate, in the Context of Quasi-Experimental Data 274
Using IVE to Resolve “Fuzziness” in a Regression-Discontinuity Design 280
What to Read Next 285
12 Dealing with Bias in Treatment Effects Estimated from Nonexperimental Data 286
Reducing Observed Bias by the Method of Stratifi cation 289
Stratifying on a Single Covariate 289
Stratifying on Covariates 299
Reducing Observed Bias by Direct Control for Covariates Using Regression Analysis 304
Reducing Observed Bias Using a Propensity-Score Approach 310
Estimation of the Treatment Effect by Stratifying on Propensity Scores 316
Estimation of the Treatment Effect by Matching on Propensity Scores 321
Estimation of the Treatment Effect by Weighting by the Inverse of the Propensity Scores 324
A Return to the Substantive Question 328
What to Read Next 331
13 Methodological Lessons from the Long Quest 332
Be Clear About Your Theory of Action 333
Learn About Culture, Rules, and Institutions in the Research Setting 335
Understand the Counterfactual 337
Always Worry About Selection Bias 338
Use Multiple Outcome Measures 340
Be on the Lookout for Longer-Term Effects 341
Develop a Plan for Examining Impacts on Subgroups 342
Interpret Your Research Results Correctly 344
Pay Attention to Anomalous Results 346
Recognize That Good Research Always Raises New Questions 348
What to Read Next 349
14 Substantive Lessons and New Questions 350
Policy Guideline 1: Lower the Cost of School Enrollment 352
Reduce Commuting Time 352
Reduce Out-of-Pocket Educational Costs 353
Reduce Opportunity Costs 354
Policy Guideline 2: Change Children’s Daily Experiences in School 355
More Books? 355
Smaller Classes? 356
Better Teaching? 357
Policy Guideline 3: Improve Incentives 359
Improve Incentives for Teachers 359
Improve Incentives for Students 361
Policy Guideline 4: Create More Schooling Options for Poor Children 364
New Private-School Options 365
New Public-School Options 365
Summing Up 367
Final Words 368
References 369
Index 381
Berminat?
Email: zanetapm@gmail.com
0 comments:
Post a Comment