Omitted Variable Bias: Missing Variables Distort Estimates
Omitted variable bias explains how missing correlated variables distort regression estimates and weaken causal interpretation in econometric models.
Causal Inference
Field Experiments Economics: Testing Policy Where People Actually Decide
Field experiments economics tests policies in real-world settings using random assignment, making causal evidence stronger while exposing implementation challenges.
Causal Diagrams DAGs: Mapping Assumptions in Research Design
Causal diagrams DAGs help researchers map assumptions, identify confounders, avoid bad controls, and design more credible empirical studies.
Propensity Score Matching: Building Comparable Groups in Observational Studies
Propensity score matching builds comparable treatment and control groups in observational studies by matching units with similar treatment probabilities.
Regression Discontinuity Design: Estimating Causal Effects at a Cutoff
Regression discontinuity design estimates causal effects at a cutoff by comparing units just below and above a policy threshold.
P-Hacking: How Hidden Choices Create False Positives
P-hacking makes fragile findings look significant by hiding specification search. Learn how the garden of forking paths affects economic research...
External Internal Validity in Economic Research
external internal validity explains how economists judge credible causal evidence inside a study and whether findings generalize across settings.
Policy Evaluation Methods: RCTs, DiD, and Synthetic Control
Policy evaluation methods like RCTs, difference-in-differences, and synthetic control build credible counterfactuals for measuring real policy effects on outcomes that...
Causal Machine Learning for Economists: Double ML and Causal Forests Explained
Causal machine learning combines flexible ML methods with econometric rigour. This guide explains Double ML and Causal Forests for economists...
Advanced Difference-in-Differences Methods: Beyond Two-Way Fixed Effects
Advanced difference-in-differences methods correct TWFE bias of 66-79% in staggered designs. Callaway-Sant'Anna, Sun-Abraham, and Goodman-Bacon have transformed applied econometrics.