Ie_Slide09 - Introductory Econometrics ECON2206/ECON3209 Slides09 Lecturer Minxian Yang ie_Slides09 my School of Economics UNSW 1 9 Specification

Info iconThis preview shows pages 1–7. Sign up to view the full content.

View Full Document Right Arrow Icon
Introductory Econometrics ECON2206/ECON3209 Slides09 Lecturer: Minxian Yang ie_Slides09 my, School of Economics, UNSW 1
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
9. Specification and Data Issues (Ch9) 9. Specification and Data Issues • Lecture plan MLR1-5 may not hold in reality. We analyse the consequences of some violations. – Functional form misspecification – Tests against nonnested models – Use proxy variables for unobserved x variables – Properties of OLS under measurement error – Missing data – Non-random samples – Outlying observations ie_Slides09 my, School of Economics, UNSW 2
Background image of page 2
9. Specification and Data Issues (Ch9) • Functional form misspecification – Misspecification • A regression is misspecified when its functional form is incorrect and fails to properly account for the relation between the dependent variable and the observable explanatory variables. • Functional form misspecification generally causes bias in estimating parameters. eg. Suppose the true model is log( wage ) = β 0 + β 1 educ + β 2 exper + β 3 exper 2 + u . Omitting exper 2 leads to biased estimation in log( wage ) = β 0 + β 1 educ + β 2 exper + v , as it misspecifies how exper affects log( wage ) . ie_Slides09 my, School of Economics, UNSW 3
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
9. Specification and Data Issues (Ch9) • Functional form misspecification – Misspecification • We consider cases where misspecification is caused by omitting a (nonlinear ) function of the explanatory variables. • We can test for this type of misspecification. – Regression specification error test (RESET) • Basic idea: when the model y = β 0 + β 1 x 1 +...+ β k x k + u is correct (ie. satisfies ZCM), additional functions of x ’s should be insignificant when added to the model. • In particular, the squared and cubed fitted values are functions of x ’s. They should be insignificant when added to the correct model. (If they are significant, then the model must be incorrect.) ie_Slides09 my, School of Economics, UNSW 4
Background image of page 4
9. Specification and Data Issues (Ch9) • Functional form misspecification – Regression specification error test (RESET) 1. OLS original model y = β 0 + β 1 x 1 +...+ β k x k + u and save the fitted values 2. Test H 0 : δ 1 = 0, δ 2 = 0 in the expanded model using the F -stat, which follow F 2, n-k -3 distribution under the null. 3. Reject H 0 when F -stat > c ( F 2, n-k -3 critical value) . ie_Slides09 my, School of Economics, UNSW 5 . ˆ y . ˆ ˆ error y δ y δ x β x β β y k k 3 2 2 1 1 1 0
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
9. Specification and Data Issues (Ch9) • Functional form misspecification – Regression specification error test (RESET) • Example 9.2. Consider the two models price = β 0 + β 1 lotsize + β 1 sqrft + β 3 bdrms + u , and log( price ) = β 0 + β 1 log (lotsize) + β 1 log(sqrft) + β 3 bdrms + v . • With
Background image of page 6
Image of page 7
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 06/12/2011 for the course ECONOMICS 3291 taught by Professor Professorsnamespublishedtheyarethesoleowners during the Three '11 term at University of New South Wales.

Page1 / 22

Ie_Slide09 - Introductory Econometrics ECON2206/ECON3209 Slides09 Lecturer Minxian Yang ie_Slides09 my School of Economics UNSW 1 9 Specification

This preview shows document pages 1 - 7. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online