Nonlinear Regression Functions Metrics Lab
Think of Diminishing Marginal Returns
Interpreting the Estimated Regression Function
What is the explanation behind? The “effect” of a change in income is greater at low than high income levels (perhaps, a declining marginal benefit of an increase in school budgets?) Other explanations…
Cubic Specification STATA
Total Production Function
Natural Logarithm
The three log regression specifications:
Logarithms in Stata Syntax: generate lnx = ln(x) generate logy = log(y) (Varname) = Expression
I. Linear-log population regression function
Linear-log case, continued
Example: TestScore vs. ln(Income)
II. Log-linear population regression function
Log-linear case, continued Standard Error of the Regression
III. Log-log population regression function
Log-log case, continued
Example: ln( TestScore) vs. ln( Income)
Summary: Logarithmic transformations
Summary: Nonlinear Regression Functions
Try it on Stata!!
Stop by the lab or email with questions Source: http://www.econ.brown.edu/fac/Frank_Kleibergen/ec163/ch08_slides_1.pdf