chapter 3. two-variable regression model: the problem of...

31
Chapter 3. Two-Variable Regression Model: The Problem of Estimation

Upload: others

Post on 07-Aug-2021

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Chapter 3.Two-Variable Regression Model:The Problem of Estimation

Page 2: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Ordinary Least Squares Method (OLS)

Recall that, PRF: Yi = β1 + β2 Xi + ui

Thus, since PRF is not directly observable, it is estimated by SRF; that is,

iii uXY ˆˆˆ21 ++= ββ

And,

iii uYY ˆˆ +=

Page 3: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

On Error Term More

iii YYu ˆˆ −=

iii uYY ˆˆ +=If

Then,

And,iii XYu 21

ˆˆˆ ββ −−=

Page 4: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

On error term moreWe need to choose SRF in such a way that, error terms should be as

small as possible,

That is,

The sum of residuals which is represented by

( )∑ ∑ −= iii YYu ˆˆ

Should be as SMALL as possible

Page 5: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

On Error Terms moreTherefore, the essential solution is to find a criterion in

order to minimize error disturbances in SRF.

All of the errors are to be as closer as possible to the

central line of SRF

Page 6: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Then, Least Squares Criterion Comes as a Solution

Least Squares Criterion is based on:

( )∑ ∑ −=22 ˆˆ iii YYu

( )∑ −−=2

21ˆˆ

ii XY ββ

Thus,

( )∑ = 212 ˆ,ˆˆ ββfui

Page 7: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Example to Least Squares Criterion

The first Model is Better?Why?

Sum of squares of Error disturbances of the second model is lower

Page 8: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Regression Equation

iii uXY ˆˆˆ21 ++= ββ

( )( )( )

( )

∑∑∑

∑∑ ∑

∑ ∑ ∑

=

−−=

−=

2i

2

222

x

ˆ

i

i

i

ii

ii

iiii

x

y

XX

YYXX

XXn

YXYXnβ

( )X

XXn

YXXYX

ii

iiiii

2

22

2

1

ˆ-Y

ˆ

β

β

=

−=

∑ ∑∑ ∑ ∑ ∑

Sample mean of YSample mean of X

Page 9: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

The Classical Linear Regression Model (CLRM): The Assumptions Underlying The Method of Least Squares

The inferences about the true β1 and β2 are important because the estimated values of them are needed to be closer and closer to population values.

Therefore CLRM, which is the cornerstone of most econometric theory, makes 10 assumptions.

Page 10: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumptions of CLRM:Assumption 1. Linear Regression Model

The regression model is linear in the parameters, that is:

Yi = β1 + β2 Xi + ui

Assumption 2. X values are fixed in repeated sampling.More technically, X is assumed to be non-stochastic

X: 80$ income level → Y: 60$ weekly consumption of a familyX: 80$ income level → Y: 75$ weekly consumption of another family

Assumption 2 is known as: Conditional Regression Analysis, that is, conditional on the given values of the regressor(s) X.

Page 11: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 3. Zero Mean value of disturbance ui

( ) 0/ =ii XuE

Page 12: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 4. Homoscedasticity or Equal Variance of ui

( ) ( )[ ]( )

cefor varian stands var

3 Assumption of because /

//var

2

2

2

where

XuE

XuEuEXu

ii

iiiii

σ=

=

−=

Page 13: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Homoscedasticity vs Heteroscedasticity

( ) 2/var σ=ii Xu

Page 14: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 5. No Autocorrelation between the disturbances

Page 15: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Autocorrelation

If :

PRF: Yt = β1 + β2Xt + ut

And if ut and ut-1 are correlated, then Yt depends not only Xt, but also on ut-1.

Page 16: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Autocorrelation in Graphs

Page 17: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 6. Zero Covariance between ui and Xi.

Page 18: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 7.

Page 19: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 8.

Page 20: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 9.

Page 21: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Assumption 10. There is No Perfect Multicollinearity

That is, there is no perfect linear relationship among the explanatory variables.

tnnt uXXXY ++++= ββββ .....22110

High correlation among independent variables causes multicollinearity which also causes standard errors to be high, hypotheses to be inefficient (low t values), etc...

Page 22: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Properties of the Least-Squares Estimators: The Gauss-Markov Theorem

Gauss-Markov Theorem is the least squares approach of Gauss (1821) with the minimum variance approach of Markov (1900).

Standard error of estimate is simply the standard deviation of the Y values about the estimated regression line and is often used as a summary measure of the “goodness of fit” of the estimated regression line.

Page 23: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

BLUE (Best Linear Unbiased Estimator)

1. An estimator is linear, that is, a linear function of a random variable, such as the dependent variable Y in the regression model.

2. An estimator is unbiased, that is, its average or expected value, E(β2), is equal to the true value, β2.

3. An estimator has minimum variance in the class of all such linear unbiased estimators; an unbiased estimator with the least variance is known as an efficient estimator.

Therefore, in the regression context it can be proved that the OLS estimators are BLUE which also sets the base of Gauss-Markov Theorem.

Page 24: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

The Coefficient of Determination, r2: A Measure of “Goodness of Fit”

The coefficient of determination, r2 (two-variable case) or R2 (multiple regression) is a summary measure that tells how well the sample regression line fits the data.

Page 25: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

The Ballentine View of R2

See Peter Kennedy, “Ballentine: A Graphical Aid for Econometrics”, Australian Economics Papers, Vol 20, 1981, 414-416. The name Ballentine is derived from the emblem of the well-known Ballantine beer with its circles.

Page 26: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Coefficient of Determination, r2

TSS = ESS + RSS

where;TSS = total sum of squaresESS = explained sum of squaresRSS = residual sum of squares

( )( ) ( )∑

∑∑∑

−+

−=

+=

2

2

2

2i ˆY

1

YY

u

YY

Y

TSSRSS

TSSESS

i

i

i

If TSS = ESS + RSS, then:

Page 27: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

On r2 more:

R2 indicates the explained part of the regression model, therefore,

TSSESSr =2

And,

( )( ) TSS

ESS

YY

YYr

i

i=

−=∑∑

2

2

Page 28: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Alternatively,

( )

TSSRSSr

YY

ur

i

i

−=

−−=∑∑

1

ˆ1

2

2

22

Page 29: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Coefficient of Determination

Page 30: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

Coefficient Of Determination

Page 31: Chapter 3. Two-Variable Regression Model: The Problem of Estimationfbemoodle.emu.edu.tr/pluginfile.php/15648/mod_resource/... · 2012. 4. 27. · Two-Variable Regression Model: The

HW # 1:

Problem 3.20 (Chapter 3)

Consumer Prices and Money Supply in Japan

1982 to 2001