Assess education–income effect credibly
Company: Google
Role: Data Scientist
Category: Analytics & Experimentation
Difficulty: medium
Interview Round: Technical Screen
Overview: Assess whether regressing income on a college-attendance dummy for 1,000 surveyed Mountain View residents recovers the effect of college on income, then design a more credible analysis. Cover confounding, covariate adjustment, instrument, cutoff and encouragement designs, and sensitivity to unmeasured confounders.
Read the full Google Data Scientist interview experience this question came from