Professional Documents
Culture Documents
Econometrics
BS Economics
Introduction
Definition of Econometrics
literally interpreted econometrics means "economic measurement". Econometrics may be defined as
the social science in which the tools of economic theory, mathematics, and statistical inference are
applied to the analysis of economic phenomena(variable).Econometrics can also be defined as
"statistical observation of theoretically formed concepts OR alternatively as mathematical economics
working with measured data. so economic theory attempts to defined the relationship among different
economic variables.
Methodology of Econometrics
Following are the main steps in methodology of econometrics
1. Specify mathematical equation to describe the relationship between economic variables.
2. Design methods and procedures based on statistical theory to obtain representative sample
from the real world.
3. Development of methods for estimating the parameters of the specified relationships
4. Development of methods of making economic forecast for policy implications based on
estimated parameters.
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
where " i " is known as the disturbance or residual term, and hence it is a random variable. It is
also
called probabilistic model. The disturbance or error term represent all those factors
affect
that
"Econometric
Model".
In other words it is an example of a "linear regression Model". In this case the response
variable 'Y' is
linearly related to the predictor variable 'X' but the relationship between the two is
second step is the estimation of model by appropriate econometric method, this will
include
the
following steps.
1. Collection of Data for the variables included in the model.
2. Choice of appropriate econometric technique for the estimation of technique used is Regression
Analysis in Statistics).
3. The third step is to develop the suitable criteria to find out whether estimates obtained are
in agreement with the expectations of the theory that has been tested, that is to decide whether
the estimates of the parameters of the theoretically meaning full and statistically significant.
4. The final step is to use the estimated model to predict the future value of the response variable.
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Types of Econometrics
Econometrics can be divided into two main braches
1. Theoretical
2. Applied
1. Theoretical econometrics:
It is concerned with development of the appropriate methods for measuring the economic relationship
specified by the econometrics models. This type of econometrics depends much on mathematical
statistics, single equation and simultaneous equations techniques, the methods used for measuring
economic relationships.
Y = a + bx simple
Y = a + bx0 + bx 2 Simul tan eous
Y = a + bx1 + bx 2 simula tan eous
2. Applied econometrics:
Applied econometrics Describe the practical value of economic research. It deals with the applications of
econometric methods developed in the theoretical econometrics to the different fields of economics
such as the consumption functions, demand and supply, fraction etc. The applied econometrics has
made it possible to obtained numerical results from these studies which are of great importance to the
planners.
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
parameters
of
the
models and finally we will carry out tests on the estimated model in an attempt to judge whether it
constitutes a sufficiently realistic picture of the economy being studied or whether somewhat different
specification is to be estimated.
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Regression Analysis
Definition of Regression
Regression analysis is statistical technique for investigating and modeling the relationship
between variables. The term regression was first time introduced by "Francis Galton". In his
paper Galton found that the height of the children of unusually tall or unusually short parents
tends to move towards the average height of the population. Galton law of universal regression
was confirmed by his friend Karl Pearson, more than a thousand records of heights of members
of family groups. He found that the average height of sons of a group of tall fathers was less
than their father height and the average height of sons of a group of tall fathers was less than
their fathers height and the average height of sons of a group of short fathers was greater
than their fathers height, thus "regressing tall and short sons alike toward the average height of
all men. In the world of Galton this was "regression to mediocrity".
Objective of Regression
The objective of regression analysis is to estimate or predict the average value of the response
variable on the basis of the known or fixed values of the predictor variable.
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
i =random error
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
An important objective of regression analysis is to estimate the unknown parameters 0 and 1 in the
regression model. This process is also called fitting the model to the data, The parameters 0 and 1 are
usually called regression coefficients. The slope 1 is the change in the mean of the distribution of Y
producing a unit change in X. The intercept 0 is the mean of the distribution of the response variable Y
when X=0.
^
^
s = ei = y i xi I
i =1
n
^
^
xi = 0
i
^
^
^
= 2 y i xi xi = 0
Yi = n + X i ii
^
Yi X i = X i + X i iii
These are called normal equations.
From equation (ii) we get:
8
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
=Y X
Now substituting the value of in equation ( ii ) we get:
^
Yi X i = (Y X ) X i + X i
Yi X i = Y X i X X i + X i
^
Yi X i Y X i = X i X X i
^
X iYi Y X i
2
Xi X Xi
X iYi Yi X i / n
2
2
X i ( X i ) / n
and
and
sample rather than the entire population. If you took a different sample, you would get different values
for
and
. Let's call
and
and
econometrics is to analyze the quality of these estimators and see under what conditions these are good
estimators and
d under which conditions they are not. Once we have
more
ore
The
variables.
second
is
The
the
first
estimates
is
of
the
the
fitted
error
terms,
and
values,
which
we
estimates
will
call
of y:
the residuals:
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
=0 +1Xi + ui
V(un)=2
10
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
5. No Autocorrelation
correlation between error term: There is no correlation between any two
error terms, symbolically:
COV( i , j )=0
11
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Interpretation of Coe
cients/Parameters
The interpretation of the coe
cients
cients from a linear regression model is fairly straightforward.
straig
The
estimated intercept (b0) tells us the value of y that is expected when x = 0. This is often not very useful,
because many of our variables dont have true 0 values (or at least not relevant ones
ones-like education or
income, which rarely if ever have 0 values). The slope (b1) is more important, because it tells us the
relationship between x and y. It is interpreted as the expected change in y for a one-unit
one
change in x.
i =1
i =1
( Xi X ) ( Y Y ) / ( Xi X )2
n
Yi (Xi--- Xi ) -- Yi
( Xi X ) /
i =1
( Xi X )2
i =1
Yi (Xi--- Xi ) /
i =1
( Xi X )2 ;
i =1
12
as
i =1
(Xi--- Xi ) = 0
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Wi
Where
xi /
Similarly we have
= y x
=
/ --(
) X
[1 / n X w ] Y
n
Hence both
and
i =1
we have
n
= Wi Yi
i =1
wi ( + Xi + i)
i =1
Now
i =1
i =1
i =1
wi = xi / xi 2
i =1
i =1
i =1
------------ I )
wi + wi xi + wi i ------------(
i =1
i =1
= ( xi x) / xi = 0 -----------(( A )
2
n
n
n
2
2
2
2
wi = xi / xi = xi /( xi ) 2 = 1 / xi ----------( B )
i =1
i =1
i =1
i =1
i =1
n
x
i
w
x
=
i i
2
i =1
i =1 x i
n
xi
( x i ) =
=1
2
x
i
13
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
n equation ( I )
Substituting these values in
^
= (0) + wi ( xi x ) + w i i
wi xi + x wi + wi i
[1 + 0] + wii
+ wii
-----------( I )
^
E
Then
Now
+ wi E ( i )
+0
since E ( i ) = 0
is an unbiased estimate of .
1 / n xw i Yi
1 / n xw ( + xi + i )
1
1
1
1
x 1
+
+ i x wi x wixi x wii
n
n
n
1
(n ) + x + 1 i x(0) x(1) x wii
n
n
1
n
+ i x wi i
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
E( )
Thus
Econometrics
BS Economics
+ E ( i ) x wi E ( i )
+0+0
1
n
since E( i ) = 0
is an unbiased estimator of .
15
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Definition
A linear regression model that involves more than one predictor variable is called
multiple linear regression model. In this case the response variable is a linear function of
two or more than two predictor variables. A multiple linear regression model with "p"
predictor variables is given by:
Yi = o + 1 X 1 + 2 X 2 + ... + p X p + i
0 , 1 ,..., p are parameters and to be estimated from the sample data. These
parameters are also called regression coefficients, the parameters
j ( j = 1,2,3... p ) represent the expected change in the response Y and percent change
in Xj, where all the remaining predictor variables Xi's (ij) are held constant. For the
reason these parameters are
Yi = 0 + 1 X 1 + 2 X 2 + i
The corresponding regression model estimated from sample data is given as:
^
Y = 0 + 1 X1 + 2 X 2
^
^
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
i =1
^
^
Yi 0 + 1 X 1 + 2 X 2
^
^
^
Yi 0 1 X 1 2 X 22 --------------( I )
e i2
^
^
The function 'S' is to be minimized with respect to 0 , 1 , and 2 . For this purpose we have to
^
^
(S )
^
( 0 )
2
ei = ^
^
0 i =1
0
(S )
^
( 1 )
^
^
^
i
0
1
1
2 X 2 = 0 -------( II )
i =1
n
^
^
^
n 2
n
e
=
Y
i
0
1
1
2 X 2 = 0 ------( III)
^
^ i
1 i =1 1 i =1
2
(S )
2
=
ei = ^
^
^
( 2 ) 2 i =1
2
17
^
^
^
i
0
1
1
2 X 2 = 0 ----( IV )
i =1
n
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
From ( II )
^
^
^
2 Yi 0 1 X 1 2 X 2 ( 1) = 0
i =1
i =1
i =1
i =1
i =1
Yi 0 1 Xi 2 X 2 = 0
i =1
i =1
Yi = 0 + 1 Xi + 2 X 2
n
i =1
Yi = n
i =1
+ 1 X 1 + 2 X 2 (V )
i =1
From ( III )
^
^
^
2 Yi 0 1 X 1 2 X 2 ( X 1 ) = 0
Yi X 1 0 X 1 1 X 1 2 X 1 X 2 = 0
^
Yi X 1 = 0 X 1 + 1 X 1 + 2 X 1 X 2 (VI )
2
^
^
^
From ( IV ) 2 Y i 0 1 X 1 2 X 2 ( X 2 ) = 0
Yi X 2 0 X 2 1 X 1 X 2 2 X 2 = 0
^
Yi X 2 = 0 X 2 + 1 X 1 X 2 + 2 X 2 (VII )
2
From ( V ) we get
18
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Yi / n = n / n
i =1
+ 1 X1 / n + 2 X 2 / n
Y = 0+ 1 X1 + 2 X 2
^
0 = Y 1 X1 2 X 2
substituting value of 0 in equation ( VI) and ( VII) we get
^
^
^
^
2
From (VI ) Yi X 1 = Y 1 X 1 2 X 2 X 1 + 1 X 1 + 2 X 1 X 2
^
^
^
^
2
From(VII ) Yi X 2 = Y 1 X 1 2 X 2 X 2 + 1 X 1 X 2 + 2 X 2
Yi X 1 Y X 1 = 1 X 1 X 1 X 1 + 2 X 1 X 2 X 2 X 1 A
^
Yi X 2 Y X 2 = 1 X 1 X 2 X 1 X 2 + 2 X 2 X 2 X 2 B
Now
)(
x1 y = X 1 X 1 Y Y
)
(
x1 y = X 1 Y Y X 1 Y Y
x1 y = X 1 Y Y , sin ce
Y Y = 0
x1 y = X 1Y Y X 1
x1 = ( X 1 X 1 ) 2
2
x1 = ( X 1 X 1 ) X 1 X 1
2
19
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
x1 = X 1 X 1 X 1 X 1 X 1 X 1
2
x1 = X 1 X 1 X 1 sin ce ( X 1 X 1 ) = 0
2
Similarly
x 2 = X 2 X 2 X 2 , sin ce ( X 2 X 2 ) = 0
2
)(
x2 y = X 2 X 2 Y Y
x2 y = X 2 Y Y X 2 Y Y
)
(
x 2 y = X 2Y Y X 2 ; sin ce Y Y
Solving the above normal equations for 0^ and 1^ i.e. multiplying equation ( C ) by
x22 an equation( d ) by x1x2 and subtract it, we will get the following estimates
of 1.
^
x1 y x 2 2 x 2 y x1 x 2
2
2
x1 x 2 ( x1 x 2 )2
1 =
Similarly multiplying equation ( C ) by ' x1x2 ' and equation ( d ) by x12 and
subtracting we will get
^
x 2 y x1 2 x1 y x1 x 2
2
2
x1 x 2 ( x1 x 2 )2
2 =
20
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Standardized coefficients:
In statistics, standardized coefficients or beta coefficients are the estimates resulting from an analysis
carried out on independent variables that have been standardized so that their variances are. Therefore,
standardized coefficients refer to how many standard
done to answer the question of which of the independent variables have a greater effect on
the dependent
variable in
a multiple
regression analysis,
when
the
variables
are
measuredindifferent unitsof (forexample, income measuredin dollars and familysize measuredin numbe
r of individuals). A regression carried out on original (unstandardized) variables produces unstandardized
coefficients. A regression carried out on standardized variables produces standardized coefficients.
Values for standardized and unstandardized coefficients can also be derived subsequent to either type
of analysis. Before solving a multiple regression problem, all variables (independent and dependent) can
be standardized. Each variable can be standardized by subtracting its mean from each of its values and
then dividing these new values by the standard deviation of the variable. Standardizing all variables in a
multiple regression yields standardized regression coefficients that show the change in the dependent
variable measured in standard deviations.
Advantages
Standard coefficients' advocates note that the coefficients ignore the independent variable's scale of units,
which makes comparisons easy.
21
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
Econometrics
BS Economics
Disadvantages
Critics voice concerns that such a standardization can be misleading; a change of one standard deviation
in one variable has no reason to be equivalent to a similar change in
another
variables are easy to affect externally, e.g., the amount of time spent on an action.
cholesterol level are more difficult, and some, like height or age, are
predictor.
Some
Weight
or
regression. The higher the coefficient of determination, the better the variance that the dependent
variable is explained by the independent variable. The coefficient of
measure of the usefulness of a regression. For example, if R2 is 0.95. This means that the variation in the
regression is 95% explained by the independent variable. That is a good regression.
Now,
if
the
Coefficient of Determination, or R2,is 0.50. Its means that the variation in the regression is 50%
explained by the independent variable. This is not a good regression. Note that R2 lies between '0' and
'1'. If R2=1, it means that the fitted model explains 100% of the variation in response variable 'Y'. On the
other hand if R2=0, the model does not explain any of
Determination can be calculated as the Regression sum of squares, RSS, divided by the total sum of
squares,SST
Coefficient of Determination =
RSS
TSS
Muhammad Ali
Lecturer in Statistics GPGC Mardan.
M.sc (Peshawar University)
Mphil(AIOU Islamabad)
^
R =
2
Econometrics
BS Economics
)(
)(
1 X1 X 1 Y Y + 2 X 2 X 2 Y Y
Yi Y
23