- Page 1 and 2: Regression‐Discontinuity Design D
- Page 3 and 4: Design Overview Design Examples Vis
- Page 5 and 6: Real Examples of RD Reading First b
- Page 7 and 8: RDD Visual Depiction Comparison Tre
- Page 9 and 10: Two Rationales for RDD 1. Selection
- Page 11 and 12: Two Rationales for RDD 1. Selection
- Page 13 and 14: 1. Overrides to the cutoff: Sharp,
- Page 15 and 16: Fuzzy RD Designs • Requires a dis
- Page 17 and 18: Requirements for Estimating Treatme
- Page 19 and 20: Implementation Threats to RD Design
- Page 21 and 22: Example: AYP Data from Texas Densit
- Page 23 and 24: What to do? • This is a big probl
- Page 25 and 26: Analytic Threats to RD Design: Miss
- Page 27 and 28: Nonlinearities in Functional Form
- Page 29 and 30: Here we see a discontinuity between
- Page 31 and 32: Functional Form: Interactions • S
- Page 33 and 34: If we superimpose the regression li
- Page 35 and 36: Here we see an example where the tr
- Page 37 and 38: Adding Nonlinear Terms to the Model
- Page 39: Adding Nonlinear and Interaction Te
- Page 43 and 44: Example of Parametric Plot used aga
- Page 45 and 46: Non‐parametric Approach lim x↓c
- Page 47 and 48: Non‐parametric plot Local linear
- Page 49 and 50: Semi‐Parametric Plot RD data from
- Page 51 and 52: What to do? • Check to see whethe
- Page 53 and 54: Shadish, Galindo, Wong, V., Steiner
- Page 55 and 56: RD Design assumption 2 • Continui
- Page 57 and 58: Implementation assumption 2 • No
- Page 59 and 60: Analytic assumption 1 continued •
- Page 61 and 62: Shadish et al. (under review) Vocab
- Page 63 and 64: RDD Example 2
- Page 65 and 66: Best case scenario - regression lin
- Page 67 and 68: Correct specification of functional
- Page 69 and 70: What to do (2) • Parametric appro
- Page 71 and 72: What to do (4) • State‐of‐art
- Page 73 and 74: What to do (5) • Move to a tie br
- Page 75 and 76: Example Cocaine Project • Include
- Page 77 and 78: What to do (7) • Estimation throu
- Page 79 and 80: 2. Dealing with Overrides to the cu
- Page 81 and 82: 2. Overrides to the cutoff • Mult
- Page 83 and 84: Specification of Functional Form Pa
- Page 85 and 86: Specification of Functional Form No
- Page 87 and 88: Overrides to cutoff (2) Alternative
- Page 89 and 90: Final estimates Function al form Pa
- Page 91 and 92:
Addressing Potential Issues with us
- Page 93 and 94:
IS Wong et al. a Fallible RD or Ins
- Page 95 and 96:
African‐American Home language Sp
- Page 97 and 98:
Fallible RD? • Now we look for di
- Page 99 and 100:
McCrary Test for New Jersey
- Page 101 and 102:
Summarizing Checks - Conditional Me
- Page 103 and 104:
Other Kinds of Threats with RD 1. L
- Page 105 and 106:
Variance of Impact Estimator in RA
- Page 107 and 108:
Variance of Impact Estimator in RDD
- Page 109 and 110:
Power Considerations Unique to RDD
- Page 111 and 112:
Generalizing Beyond The Cut‐off I
- Page 113 and 114:
Observed vs Counterfactual Outcomes
- Page 115 and 116:
Questions About Generalizing From R
- Page 117 and 118:
What Do We Mean By Comparison Group
- Page 119 and 120:
Our Purpose • Use experimental da
- Page 121 and 122:
What Is The Target Parameter? • A
- Page 127 and 128:
What About The Above The Cut‐Off
- Page 129 and 130:
Root Mean Square Error In Experimen
- Page 131 and 132:
Post-Treatment Medicaid Expenditure
- Page 133 and 134:
Results For Number of Unpaid Worker
- Page 135 and 136:
Results For Expenditures With Expen
- Page 137 and 138:
Conclusions re Using a Comparison R
- Page 139 and 140:
Implications for Practice • Which
- Page 141 and 142:
RD with Multiple Cutoffs Design (as
- Page 143 and 144:
Analyzing RDs with Multiple Cutoffs
- Page 145 and 146:
Examples • Black, Galdo, and Smit
- Page 147 and 148:
Issue 2: To pool or not pool across
- Page 149 and 150:
Pooling is Already Common in Social
- Page 151 and 152:
What is Special about Pooling RD Es
- Page 153 and 154:
Estimating Treatment Effects across
- Page 155 and 156:
Estimating Treatment Effects across
- Page 157 and 158:
Generalizing Across Multiple Assign
- Page 159 and 160:
Multivariate RDD with Two Assignmen
- Page 161 and 162:
Recent Education Examples of RDDs w
- Page 163 and 164:
Frontier‐specific Effect (τ R )
- Page 165 and 166:
Requirements for a valid Multivaria
- Page 167 and 168:
Frontier Approach Estimates the dis
- Page 169 and 170:
Univariate Approach Addresses dimen
- Page 171 and 172:
Monte Carlo Study Wong, V., Steiner
- Page 173:
Implications for Practice • Which