ECO375H_Slides_2

# ECO375H_Slides_2 - Lecture 2 The Simple Regression Model...

This preview shows pages 1–16. Sign up to view the full content.

Lecture 2: The Simple Regression Model Junichi Suzuki University of Toronto September 17th, 2009

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

View Full Document
Announcement I Problem set 1 is posted online and due September 25th I Two normal questions and two STATA questions I On Friday, Sacha will present the very basic of STATA I On 23rd (Wed), Sacha will have an o¢ ce hour to help students having STATA-related problems
Some Advice on Problem Set 1 I Start early! Need time to get used to use software! I I Will post some hints on 2.10 in the blackboard on Friday

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

View Full Document
Chapter 2 The Simple Regression Model
I Provide a quick review of the ordinary least squares estimates (OLSE) using cross-sectional data I Theoretical aspects: I De±nition I How to derive I Mechanical and stochastic properties and assumptions needed I Practical aspects: I How to interpret results I How to deal with nonlinearity

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

View Full Document
Model
I Want to quantify the change of x on y with all other factors being equal (e.g., x : training and y : wage) I Suppose all other factors are not observable I One relevant econometric model may be a simple linear regression model: y = β 0 + β 1 x + u I y : dependent variable/regressend I x : independent variable/regressor I u : error term/disturbance I β 0 : intercept parameter/constant term I β 1 : slope parameter

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

View Full Document
Assumptions I The average value of u is zero E ( u ) = 0 I Not restrictive I When E ( u ) 6 = 0 , β 0 and u so that ˜ β 0 = β + E ( u ) and E ( ˜ u ) = 0 I The average value of u does not depend on that of x E ( u j x ) = E ( u ) I Restrictive! I Imply that workers±tendency to attend training does not depend on their IQ I The most important assumption in econometrics
Population Regression Function (PRF) I Under these two assumptions, we can separate the model in two parts: y = β 0 + β 1 x + u = E ( y j x ) + u I E ( y j x ) = β 0 + β 1 x : PRF or systematic part I u : unsystematic part I PRF tells how the average value of y changes as the value of x changes

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

View Full Document
Figure 2.1 should come here
2.2: Deriving the OLSE

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

View Full Document
Deriving the OLSE I Consider a simple regression model: y = β 0 + β 1 x + u I Suppose you observe n data points: f y i , x i g n i = 1 I β 0 and β 1 that I OLSE is the most popular method for its simplicity and good properties
Deriving the OLSE I ˆ β 0 , ˆ β 1 ± that solve min β 0 , β 1 n i = 1 ² y i β 0 β 1 x i | {z } ³ 2 residual I minimizes the sum of squared "residual" (SSR) I This is a mere mathematical problem I Solving this problem entails the knowledge of basic calculus

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

View Full Document
Figure 2.4 should come here
Deriving the OLSE I At the minimum, its partial derivative w.r.t. β 0 , β 1 must be zero I FOCs are ∂β 0 n i = 1 ( y i β 0 β 1

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

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 10/05/2009 for the course DEPARTMENT Eco375 taught by Professor Suzuki during the Spring '09 term at University of Toronto.

### Page1 / 46

ECO375H_Slides_2 - Lecture 2 The Simple Regression Model...

This preview shows document pages 1 - 16. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online