Table 6.

Meta-Regression of Standardized Absolute Bias.

	Coefficient	95%CI
Constant	0.13	−0.07	0.32
Regression estimation*
Regression discontinuity design	−0.12	−0.31	0.07
Matching	−0.16	−0.36	0.04
No baseline measure*
Baseline measure	−0.01	−0.06	0.03
Non-local comparison*
Local comparisons	−0.01	−0.06	0.03
Parsimonious controls*
Rich control variables	−0.05	−0.09	−0.01
Nearest-neighbor matching	−0.02	−0.10	0.06
Kernel matching	0.12	−0.06	0.30
Other matching algorithm*
Education outcome*
Health outcome	0.08	−0.03	0.18
Income outcome	0.14	0.02	0.27
Some concerns about bias*
High risk of bias	0.07	0.01	0.13
Number of obs	586
Tau-squared	0.00
I-squared	0.0%
F	2.48
Prob > F	0.01

Notes: * base category. Standard errors use cluster adjustment (equation (5)).