The Simple Linear
Regression Model
Ordinary Least Square Method
Simple linear regression model
e.g : Supply Function
Y = f(X)
Where Y = quantity supplied
X = price of the commodity
Assuming that variables are related
with the simplest possible
mathematical form, so the supply
function in linear form is
Yi = b0 +b1Xi
b0 and b1 are the parameters of supply
function and should be estimated its
numerical value, o and 1
o
should be either zero or positive
0)
In the particular case of supply
function, the sign of 1 expected to be
positif
( 1 > 0)
It is important to examine the
relationship between the price
elasticity of supply and the coefficient ,
0 and
1
Elasticity from a regression line, use
estimated and the mean value of
price ( )
and quantity ( )
Remember:
so, the form would be
Given that
, it follows that:
(i) The supply will be elastic (
) if
is negative (
)
(ii) The supply will be inelastic (
) if
is positive (
)
(iii)The supply with have unitary elastic if
The above form implies that the relationship
between X and Y is exact, there is no other
factors that is effecting Y.
But, when we gather the observation about
the quantity supplied in the market with
various price, well have the diagram
e. g :
The deviation of the observation from line may be
attributed to several factors
1. Omission of variables from the function
2. Random behaviour of the human being
3. Imperfect spesification of the mathematical from
the model
4. Errors of aggregation
5. Error of measurement
Error in econometic function is usually donated by the
letter u and is called error term or random
disturbance term or stochastic term
The function model would be
The true relationship which connects the vaiables involved is
splite into two parts :
A part represented by a line
A part represented by the random term u
Look the figure
(figure shows the term d
refers to term u)
Variation
in Y
Systemati
+c
Variation
Variation
in Y
Explaine
d
=
Variation
Random
Variatio
n
Unexplained
Variation+
To estimate the coefficient b0 and b1 we
need observation on X, Y and u. Yet, u is
never like other explanatory variables, and
therefore in order to estimate the function
we should guess the values of u, that is we
should make some reasonable assumptions
of each ui (its mean, variance and
covariance)
Assumption of the linear stochastic regression model
Stochastic assumption of ordinary lest
square
Assumption 1 (randomness of u)
ui is a random real variable : it may bepositif, negative or zero
Assumption 2 (zero mean of u)
The mean value of u in any particular period is zero
Assumption 3 (homoscedasticity)
The variance of ui is constant in each period
Assumption 4
the variable ui has a normal distribution
Normal Distribution
Assumption 5 (Nonautocorrelation)
The random term of different observation (ui, uj) are
independent: covariance of any ui with any other uj are equal to
zero
for
Assumption 6
u is independent of explanatory variable(s) : their covariance is
zero
Assumption 6A
The Xis are set of fixed value in the hypothetical process of
repeated sampling which underlies the linear regression model
Assumption 7 (No errors of measurement in the Xs)
The explanatory variable(s) are measured without error
Other assumption of ordinary least
square
Assumption 8 (Noperfect multicolinear Xs)
The explanatory variables are not perfectly
linearly correlated
Assumption 9
The macrovariables should be correctly
aggregated
Assumption 10
The relationship being estimated is identified
Assumption 11
The relationship is correctly specified
The distribution of the dependent
variable Y
Dependent variable Y has a normal distribution with mean
And variance
Proof 1.
Given
Taking expected values we find
Bu using assumption 6
Furthermore, by assumption 2
Therefore,
Proof 2.
By assumption 3, the uis are
homoscedastic, that is, they have the
constant variance
Therefore,
The least square criterion and the
normal equation of OLS
The true relationship between X and Y is
The true regression line is
And the estimated relationship is
And the estimated regression line is
Where = estimated value of Y, given a specified
of X
=estimate of the true intercept
= estimate the true parameter
= estimate of th true value of the random term u
value
the striped line shows estimated regression
line
And the light line shows the true regression
line
Clearly deviation of the observation from the lines depend on their
constant intercept (
) and their slope (
). The choice among all
possible lines is done on the basis of what is called the least squares
criterion.
the least squate method should now be clear: the method seeks
the minimisation of the sum of the squares of the deviation of
the actual observation from the line
To estimate the function by calculating the
and
value, we can use these form
Or by using the deviation of the variables
from the data mean
Worksheet for the estimation supply
n
Yi
Xi
69
76
12
52
56
10
57
77
10
58
Xi2
XiYi
yi
xi
xiyi
xi2
81
621
144
912
13
39
36
312
-11
-3
33
100
560
-7
-7
81
513
-6
100
770
14
14
49
406
-5
-2
10
55
64
440
-8
-1
67
12
144
804
12
10
53
36
318
-10
-3
30
11
72
11
121
792
18
12
64
64
512
-1
-1
Estimation of a function whose
intercept is zero
e.g : linear production function of
manufactured products should normally
have zero intercept
Estimated function
Imposing the restriction
Therefore,
Estimation of elasticities from an
estimated regression line
Estimated function
The derivative of
with respect to X
If the estimated function is a linear demand of supply function the
coefficient
is not the price elasticity, but a component of the
elasticity, which defined by
Where
= price elasticity
Y = quantity
X = price
Clearly
is the component
From the estimated function we obtain an average
elasticity
Where = the average price
= average regressed value of the
quantity, i.e the mean value of the
estimated fromthe regression s
= average value of the quantity
Note that
, that is,the mean of the estimated
value of Y is equal to the mean of the actual
(sample) values of Y, because
