B
Brilly's Learning Plan
DashboardRoadmapEvidenceProfileSettings
B
Brilly's Learning Plan
DashboardRoadmapEvidenceProfileSettings
Back to Roadmap
Not StartedMedium Priority

Panel / longitudinal regression basics

Why It Matters

I need to know what claims I can make. 'Association' is easy; 'Causality' is hard. Panel data brings me closer to causality than cross-sectional.

What To Learn

  • Fixed vs Random Effects: Controlling for unobserved heterogeneity (the 'personalality' of the student/university).
  • Lagged Variables: Using X at time T-1 to predict Y at time T.
  • Standard Errors: Clustering them by university/classroom to avoid false confidence.

Tags

Quantitative methodsMonth 4-6

History