Linear in Parameters

A model that is linear in the β\betas after the variables are transformed (logs, squares), so OLS still applies even if the relation between xx and yy is curved.

When to use

Check this before using OLS on a nonlinear story: logs and quadratics are fine, but if no transformation makes the model linear in its parameters, OLS fails and you would need NLLS or GMM.

See Lec 3.

Appears in

No references yet.