2. Econometrics is the application of statistical,
and mathematical techniques to analysis
economic data with a purpose of testing and
measuring different theories.
Micro Economic Theory
Macro Economic Theory
We have to test with the actual data.
What is the magnitude and sign?
2
3. 1. Statement of theory or hypothesis.
what we want to test?
Keynes theory there is posetive relation ship between income
and consumption, for a unit increment of income say of $1, the
increment in consumption is greater than zero but less than
one).
2. Specification of the mathematical model
Where: 0 < β < 1
3
X
Yi
4. In mathematical model for fixed value of income
there is fixed level of consumption
The relationship between income and
consumption is exact or deterministic
Income is the only determinants of consumption.
It is impossible/ very difficult to estimate the
values of parameters
3. Specification of econometric model
relationships between economic variables are
generally inexact.
4
5. To allow for the inexact relationships between economic
variables, the econometrician modify the deterministic
model as.
Consumption is a function of income and Error term
Why do we need to include the stochastic (Error term)
component, for example in the consumption function?
5
i
i u
X
Y
6. Omission of variables
The functional form may not be correct.
randomness on human behavior
measurement error in collecting data.
4. Obtaining Data
Cross sectional Data
Time series Data
Pooled Data
Panel Data
6
7. 5.Estimation of the parameters
Ordinary least square method (OLS)
Maximum likelihood method (MLM)
Method of moments (MM)
7
8. 6. Evaluation of parameters
Economic or prior criteria
Statistical criteria
Econometric criteria
7. Forecasting and controlling of policy
What will be the level of consumption if income is
3000
If government targets to reach the consumption of
households to be 2000 income has to be?
8
9. Regression is estimation or prediction of the
average value of a dependent variable on the
basis of the fixed values of independent
variables .
9
11. unconditional mean is $121.20 ($7272/60)
Increment in income led to increment in
average consumption of households but not
necessarily for individual household
consumptions.
a population regression curve (Function) is
simply the locus of the conditional means of
the dependent variable for the fixed values of
the explanatory variable(s).
11
13. Regression is estimation or prediction of the average
value of a dependent variable on the basis of the
fixed values of other variables.
Correlation measures the strength of linear
association between variables.
In regression, we have stochastic dependent variable
and non-stochastic independent variable (fixed)
while in correlation, variables involved are
stochastic.
Causation comes from theory rather than statistics.
Thus, regression and correlation does not necessarily
imply causation.
13
15. 15
In regression analysis our interest is estimating
the PRFs that is, estimating the values of the
unknowns βo and β1 on the basis of
population observations of Y and X.
But in most practical situations what we have
is a sample of Y values corresponding to some
fixed X’s. Therefore, our task now is to
estimate the PRF on the basis of the sample
information.
33. If our objective is to estimate the sample
parameter only the method of OLS discussed
in the preceding section will be enough.
But our objective is not only to obtain the
estimates but also to draw inferences about
the true parameter.
How the sample parameter and conditional
mean close to their population counter part.
33
39. Confidence Interval
This interval is called confidence interval.
1-α is called the confidence coefficient.
α is called the level of significance
(probability of committing Type I error)
The larger is the standard error, the larger is
the width of the confidence interval.
39