Methods in Psychology · 2021 · 147 citations · 30 references
Public HealthEconomicsCausal InferenceEconometricsExperimental DesignBiasBias ControlCausal Effect EstimationInstrumental VariablesRegression Discontinuity DesignsSelection BiasBias ReductionCausal RelationshipsInstrumental Variable AnalysisInstrumental Variables EstimationInstrumental Variables RegressionTwo-stage Least SquaresTwo-stage Least Squares RegressionUnobserved Heterogeneity
We aim to raise awareness of the omitted variable bias (i.e., one special form of endogeneity) and highlight its severity for causal claims. Firstly, we demonstrate via analytic proof that omitting a relevant variable from a model which explains the independent and dependent variable leads to biased estimates. Secondly, we offer an easy-to-understand visualization for the problem. Finally, we discuss two remedies, diminishing the risk of the omitted variable bias, namely the instrument variable or two-stage least squares estimator and the regression discontinuity design. We hope that our review will motivate researchers to use them more often in future research.
30
Biometrika · 1908 · 3.9K citations
John Bound, David A. Jaeger, Regina Baker · Journal of the American Statistical Association · 1995 · 3.7K citations
The Estimation of Economic Relationships using Instrumental Variables
J. D. Sargan · Econometrica · 1958 · 3.4K citations