ECON206_1011_01_Handout_10

# ECON206_1011_01_Handout_10 - ECON 206 METU Department of...

This preview shows pages 1–5. Sign up to view the full content.

ECON 206 December 27, 2010 METU- Department of Economics Instructor: H. Ozan ERUYGUR e-mail: [email protected] 1 LECTURE 10 SIMPLE REGRESSION MODEL - II Outline of today’s lecture: I. Mean and Variance of the Dependent Variable Y . ............................................................. 1 II. Ordinary Least Squares (OLS) Estimation. ....................................................................... 2 A. Mean of 1 ˆ β .................................................................................................................... 6 B. Variance of 1 ˆ ............................................................................................................... 9 C. Mean of 0 ± ................................................................................................................. 10 D. Variance of 0 ± ............................................................................................................ 11 E. Covariance of 0 ˆ and 1 ˆ ............................................................................................. 13 I. Mean and Variance of the Dependent Variable Y The dependent variable Y has mean 01 () tt EY X = + and variance 2 22 ( ) t t Var Y E Y E Y E u σ ⎡⎤ ⎣⎦ =− = = 1. Let us show that the mean of Y t is t X + . Æ By definition the mean of Y t is its expected value. Given that t YX u =+ + . Taking the expected values we get ) ) ( ( t t E X u E X Eu ββ + = ++

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

View Full Document
ECON 206 December 27, 2010 METU- Department of Economics Instructor: H. Ozan ERUYGUR e-mail: [email protected] 2 Given that 0 β and 1 are parameters and by Assumption 2 the values of X t ’s are a set of fixed numbers (in the process of hypothetical sampling) 01 tt EX X ββ ⎡⎤ = ⎣⎦ + + Furthermore, by Assumption 3, ()0 t Eu = . Therefore, ) ( EY X = + 2. Let us now show that the variance of Y t is 2 σ 2 () t Var Y E Y E Y =− 2 t t Var Y E X u X =+ + + 2 Var Y E u = By Assumption 4, the u t ’s are homoscedastic, that is, they have the constant variance 2 2 2 Var Y E u = = II. Ordinary Least Squares (OLS) Estimation The two-variable population regression function is given by t YX u + , but we do not observe it. Hence we estimate it from the sample regression function
ECON 206 December 27, 2010 METU- Department of Economics Instructor: H. Ozan ERUYGUR e-mail: [email protected] 3 01 ˆ ˆˆ ˆ t tt t Y YX u ββ =+ + ±²³²´ . or ˆ ˆ ttt YYu . We can rewrite the sample regression function as ˆ ˆ tttt t uY YY X β =−=− . In other words, the residuals are the differences between the actual and the estimated t Y values. With T observations, we want to choose 0 ˆ and 1 ˆ such that the sum of the residuals is minimized: 11 ˆ ˆ () . TT t Y == =− ∑∑ this turns out not to be a very good rule because some residuals are negative and some are positive (and they would cancel each other), and

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

View Full Document
ECON 206 December 27, 2010 METU- Department of Economics Instructor: H. Ozan ERUYGUR e-mail: [email protected] 4 all residuals have the same weight (importance) even though some are small and some are large. 9 To overcome these problems, we use the squares of the residuals instead of their own values. Ordinary Least Squares (OLS) criterion: Minimize 2 2 2 01 ˆˆ ˆ ˆ () ( ) tt t t t uY Y Y X ββ =− ∑∑ wrt 1 ˆ β and 2 ˆ The necessary condition for a minimum is that the first derivatives of the function be equal to zero. Partial differentiation yields 2 0 ˆ 2( ) 0 ˆ t u YX = ( 1 ) 2 1 ˆ ) 0 ˆ t t u X = ( 2 ) From (1) 0 −− = 11 . 0
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

### Page1 / 14

ECON206_1011_01_Handout_10 - ECON 206 METU Department of...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online