Presentation is loading. Please wait.

Presentation is loading. Please wait.

ECONOMETRICS DR. DEEPTI.

Similar presentations


Presentation on theme: "ECONOMETRICS DR. DEEPTI."— Presentation transcript:

1 ECONOMETRICS DR. DEEPTI

2 What is Regression Study of relationship between the explained or dependent variable and one or more independent or explaining variables Relationship does not imply causation Is a conditional mean, i.e. if we are given the value of certain independent variables (Xi’s) E(Y/Xi) May be conducted for the following reasons :- To find the conditional mean of Y, given X To test the relationship between X and Y To predict the value of Y for a given X

3 Population Regression Function (PRF)
Systematic Determination Component Random Component

4 What does error term represent
Sample Regression Function (SRF) What does error term represent Effect of variables not included in the model Errors of measurment

5 What is Linear Regression
Why use a method Since we would like to get the estimate of Y given the Xi Plot a line that minimizes Ui’s Since some Ui > 0 and some Ui <0 Several lines such that Hence two options - minimized What is Linear Regression Linear in variables linear in parameter

6 The method of ordinary Least Squares
The two variable PRF : We estimate from the SRF :

7 Least Square estimation method :-
When the actual values of X and Y are used When the values are taken as the derivation from the actual mean Squaring and summing on both sides we obtain It has to be minimized FOC

8 Matrix notation of Eq. (1) and Eq. (2)
By Cramer rule

9 And from Eq. (1) SOC Second order condition of minimization is that H>0, and second order partial differentiation must be positive. (i) (ii)

10 Assumptions of (Ui) Linear Regression Model
Ui is a random real variable and has normal distributions The mean value of ui is zero E(Ui)=0 {i=1,2,3,……….n} The variance of ui is constant E(Ui2)=2 This assumption is known as the assumption of Homoscedasticity The disturbance terms of different observation (UiUj) are independent E(UiUj)=0 (i≠j) The assumption is known as the assumption of Non autocorrelation The explanatory variables is non-stochastic variable and is measure without error, Ui is independent of the explanatory variables. E(XiUj)= Xi E(Uj)=0 (for all j=1,2,3,………n)

11 Least Square estimation of
Standard error of estimate

12 Coefficient of determination r2
Is a measure of goodness of fitness Measures what percentage of deviation of Y from its mean, is explained by the deviation of X from its mean

13 Coefficient of Correlation (R)
A measure of strength of linear relationship between two variables. Is the square root of the coefficient of determination R takes the same sign as the slope coefficient,

14 Functional forms of Regression Models
Semi log Model How to measure the growth rate – The log – lin model R= compound rate of growth of Y This model is like any other linear regression model in that the parameters and are linear

15 THANK YOU


Download ppt "ECONOMETRICS DR. DEEPTI."

Similar presentations


Ads by Google