What are assumptions of Linear Regression?

Updated May 16, 2026

Short answer

Linearity, independence, homoscedasticity, normality, and no multicollinearity.

Deep explanation

These assumptions ensure valid statistical inference. Violations can lead to biased or inefficient estimates.

Real-world example

Financial forecasting models rely heavily on these assumptions.

Common mistakes

  • Ignoring residual diagnostics.

Follow-up questions

  • How do you test homoscedasticity?

More Linear Regression interview questions

View all →