This preview shows pages 1–6. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: 1 / 28 Introduction to Econometrics Econ 322 Section 1 Fall, 2010 Lecture 19: Nonlinear models (cont) November 8, 2010 Topics Covered triangleright Topics Covered Logarithmic functions of Y and/or X Three Important Cases I. LinearLog Model II. LogLinear Model LogLog Model Comparison of Loglinear and LogLog Summary: Logarithmic transformations Interactions between Variables Interactions between two binary variables Our Earnings Example Interactions between Binary and Continuous Variables Interactions between two continuous Variables 2 / 28 1. logarithmic functions 2. modelling interactions between variable Logarithmic functions of Y and/or X Topics Covered triangleright Logarithmic functions of Y and/or X Three Important Cases I. LinearLog Model II. LogLinear Model LogLog Model Comparison of Loglinear and LogLog Summary: Logarithmic transformations Interactions between Variables Interactions between two binary variables Our Earnings Example Interactions between Binary and Continuous Variables Interactions between two continuous Variables 3 / 28 square Let ln(X) = the natural logarithm of X square Logarithmic transforms permit modeling relations in percentage terms (like elasticities), rather than linearly. square Heres why: square (calculus: dln ( x ) dx = 1 x ) square Numerically: ln (1 . 01) = . 00995 = . 01; ln (1 . 10) = . 0953 = . 10 (sort of) Three Important Cases Topics Covered Logarithmic functions of Y and/or X triangleright Three Important Cases I. LinearLog Model II. LogLinear Model LogLog Model Comparison of Loglinear and LogLog Summary: Logarithmic transformations Interactions between Variables Interactions between two binary variables Our Earnings Example Interactions between Binary and Continuous Variables Interactions between two continuous Variables 4 / 28 Case Population Regression Function I. LinearLog y i = + 1 log( x i ) + epsilon1 i II. LogLinear log( y i ) = + 1 x i + epsilon1 i III. LogLog log( y i ) = + 1 log( x i ) + epsilon1 i square The interpretation of the slope coefficient differs in each case. square The interpretation is found by applying the general before and after rule: figure out the change in Y for a given change in X. I. LinearLog Model Topics Covered Logarithmic functions of Y and/or X Three Important Cases triangleright I. LinearLog Model II. LogLinear Model LogLog Model Comparison of Loglinear and LogLog Summary: Logarithmic transformations Interactions between Variables Interactions between two binary variables Our Earnings Example Interactions between Binary and Continuous Variables Interactions between two continuous Variables 5 / 28 square interpretation of 1 consider what happens when we change X by a small amount 1. before: 2. after: difference is y = 1 [log( x + x ) log( x )] = 1 x x ....
View Full
Document
 Fall '11
 LANDONLANE
 Econometrics

Click to edit the document details