Regression Foundations

Routing Summary

This folder covers regression mechanics and interpretation from MHE Chapter 3. Contains 3 notes.

Concept Map

ConceptNoteTypeDepends OnKey Result
Regression as best linear approximation to CEF, Frisch-Waugh, robust SEsRegression and the CEFconceptThe Selection Problem, The Experimental Ideal, Research Questions in EconometricsRegression is the minimum MSE linear approximation to the CEF
CIA / selection on observables: when regression is causalConditional Independence AssumptionconceptRegression and the CEF, The Selection Problem, The Experimental IdealCIA + overlap regression estimates causal effects
OVB formula, direction of bias, schooling exampleOmitted Variables BiasconceptRegression and the CEF, Conditional Independence Assumption, The Selection ProblemOVB = (effect of omitted var) x (relationship to included var)

Notes

  • Regression and the CEF — CONTAINS: Conditional expectation function, regression as CEF approximation, Frisch-Waugh theorem, robust standard errors, saturated models
  • Conditional Independence Assumption — CONTAINS: CIA / selection on observables, overlap condition, propensity score, matching, when regression is causal
  • Omitted Variables Bias — CONTAINS: OVB formula, short vs long regression, direction of bias, schooling returns example, bad controls

Sources

See Also