HARAMAYA UNIVERSITY

COLLEGE OF AGRICULTURE AND

ENVIRONMENTAL SCIENCES

Econometrics
Kindineh Sisay
(Lecturer of Agricultural and Applied Economics)
Email: [email protected]

January, 2022
Haramaya University
 CHAPTER OUTLINES
Unit 1: Fundamental concepts of Econometrics
Unit 2: Correlation Theory
Unit 3: Simple Linear Regression Models
Unit 4: Multiple Regression Analysis
Unit 5: Econometric Problems
Unit 6: Non-linear regression and Time series
Economtrics
 CHAPTER ONE
INTRODUCTION TO ECONOMETRICS

Outlines:
Definition and scope of econometrics

Why a Separate Discipline?

Economic models vs. econometric models

Methodology of econometrics

Desirable properties of an econometric model

Goals of econometrics
 DEFINITION AND SCOPE

What is Econometrics?

oIt is a social science in which the tools of economic theory,

mathematics and statistical inference are applied to the analysis
of economic phenomena.
o Econometrics is the science which integrates economic theory,
economic statistics, and mathematical economics to investigate
the empirical support of the general law established by
economic theory.
oIt is concerned with the measuring of economic relationships.
 CONT’D…

In short, econometrics may be considered as the integration

of economics, mathematics, and statistics for the purpose of
providing numerical values for the parameters of economic
relationships and verifying economic theories.
ECONOMETRICS VS. MATHEMATICAL ECONOMICS & ECONOMIC THEORY

Mathematical economics states economic theory in terms of

mathematical symbols while economic theory use verbal
exposition.
 There is no essential difference between them.
 Both express economic relationships in an exact or deterministic
form.
 Neither mathematical economics nor economic theory allows
for random elements which might affect the relationship and
make it stochastic.
 CONT’D…
 Further, they do not provide numerical values for the coefficients
of economic relationships.
 Econometrics differs from mathematical economics in that, it
assumes random relationships among economic variables.
 Econometric methods are designed to take into account random
disturbances which relate deviations from exact behavioral
patterns.
Moreover, econometric methods provide numerical values of the
coefficients of economic relationships.
 Example: Law of Demand ...
 CONT’D…
Econometrics vs. statistics

 An economic statistician gathers empirical data, records them,

tabulates them or charts them, and attempts to describe the
pattern in their development over time
 Economic statistics is mainly a descriptive aspect of
economics.
 It does not provide explanations of the development of the
various variables and it does not provide measurements for the
coefficients of economic relationships.
 CONT’D…

 Econometrics utilizes statistical data to estimate

quantitative economic relationships and to test hypothesis

about them.
 CONT’D…
Economic models vs. Econometric models

What is model for you?

A model is an organized set of relationships that describes the
functioning of an economic entity
It is a simplified representation of a real-world process under
a set of simplifying assumptions.
The real world system is represented by the model in order to
explain it, to predict it, and to control it.
 CONT’D…

 A given representation of real world system can be a model if it

fulfills the following requirements.

 A good model is both realistic and manageable.

 CONT’D…
Economic models

 Any economic theory is an observation from the real world.

 It is an organized set of relationships that describes the
functioning of an economic entity under a set of simplifying
assumptions.
 All economic reasoning is ultimately based on models.
 CONT’D…
Economic models consist of the following three basic structural
elements.

 A set of variables
 A list of fundamental relationships and
 A number of strategic coefficients
 CONT’D…
Econometric models
 They contain a random element which is ignored by
mathematical models and economic theory which postulate
exact relationships between economic variables.
 Example: Economic theory postulates that the demand for a
commodity depends on its price, on the prices of other related
commodities, on consumers’ income and on tastes.
 This is an exact relationship which can be written
mathematically as:
 CONT’D…
 The above demand equation is exact.
 However, many more factors may affect demand.
 In econometrics the influence of these ‘other’ factors is taken into
account by introducing random variable.

 The random term (also called error term or disturbance term) is a

surrogate variable for important variables excluded from the
model, errors committed and measurement errors.

+ Ui
 METHODOLOGY OF ECONOMETRICS

 Starting with the postulated theoretical relationships

among economic variables, econometric research or inquiry
generally proceeds along the following lines/stages:

 Specification of the model

 Estimation of the model
 Evaluation of the estimates
 Evaluation of he forecasting power of the estimated
model
 CONT’D…
1. Specification of the model
In this step the econometrician has to express the relationships between
economic variables in mathematical form.

The step involves the determination of three important issues:

i. Determine dependent and independent (explanatory) variables to be included

in the model,

ii. Determine a priori theoretical expectations about the size and sign of the
parameters of the function, and

iii. Determine mathematical form of the model (number of equations, specific

form of the equations, etc.).
 CONT’D…
 Specification of the econometric model will be based on economic theory
and any available information related to the phenomena under investigation.

 Moreover, knowledge of economic theory and familiarity with the

particular phenomenon being studied is also needed.

 It is the most important and the most difficult stage of any econometric
research.
 CONT’D…
likelihood of committing errors or incorrectly specifying the model.

The most common errors of specification are:

 Omissions of some important variables from the function.

 The omissions of some equations (for example, in simultaneous equations

model).

 The mistaken mathematical form of the functions.

 CONT’D…
2. Estimation of the model

This is purely a technical stage which requires knowledge of the various

econometric methods, their assumptions and the economic implications for
the estimates of the parameters.

This stage includes the following activities:

 Gathering of the data on the variables included in the model.

 Examination of the identification conditions of the function (especially for

simultaneous equations models).
 CONT’D…
 Examination of the aggregations problems involved in the variables of the
function.

 Examination of the degree of correlation between the explanatory variables

(i.e., examination of the problem of multicollinearity).

 Choice of appropriate economic techniques for estimation, i.e. to decide a

specific econometric method to be applied in estimation; such as, OLS,
MLM…
 CONT’D…
3. Evaluation of the estimates

 Consists of deciding whether the estimates of the parameters are theoretically

meaningful and statistically significant.

 Enables the econometrician to evaluate the results of calculations and

determine the reliability of the results.

Criteria:

Economic a priori criteria: determined by economic theory and refer to the

size and sign of the parameters of economic relationships.
 CONT’D…
 Statistical criteria (first-order tests):
 These are determined by statistical theory and aim at the evaluation of the
statistical reliability of the estimates of the parameters of the model.

o Correlation coefficient test, standard error test, t-test, F-test, and R 2-test
are some of the most commonly used statistical tests.

Econometric criteria (second-order tests): These are set by the theory of

econometrics and aim at the investigation of whether the assumptions of
the econometric method employed are satisfied or not in any particular
case.
 CONT’D…
 The econometric criteria serve as a second order test (as test of the
statistical tests)

 i.e., they determine the reliability of the statistical criteria;

 they help us establish whether the estimates have the desirable
properties of unbiasedness, consistency, etc.

 Econometric criteria aim at the detection of the violation or validity of the

assumptions of the various econometric techniques.
 CONT’D…
4. Evaluation of the forecasting power of the model

 Forecasting is one of the aims of econometric research.

 However, before using an estimated model for forecasting by
some way or another, the predictive power and other
requirements of the model need to be checked.
 It is possible that the model may be economically meaningful
and statistically, and econometrically correct for the sample
period for which the model has been estimated.
 CONT’D…

 This stage involves the investigation of the stability of the

estimates and their sensitivity to the changes in the size of the
sample.

 The estimated function should performs adequately outside the

sample data which require model performance test under extra
sample.
DESIRABLE PROPERTIES OF AN ECONOMETRIC
MODEL

The ‘goodness’ of an econometric model is judged

customarily based on the following desirable properties:

 Theoretical plausibility;

 Compatibility: The model should be compatible with the postulates

of economic theory and adequately describe the economic
phenomena to which it relates.
 CONT’D…
 Explanatory ability:
The model should be able to explain the observations of the
actual world.
It must be consistent with the observed behavior of the economic
variables whose relationship it determines.
 Accuracy of the estimates of the parameter:
The estimates should possess (if possible) the desirable
properties of unbiasedness, consistency and efficiency.
 CONT’D…
 Forecasting ability: The model should produce satisfactory
predictions of future values of the dependent variables.

 Simplicity: The model should represent the economic

relationships with maximum simplicity.

The fewer the equations and the simpler their mathematical form, the
better the model provided that the other desirable properties are not
affected by the simplifications of the model.
 GOALS OF ECONOMETRICS

Basically there are three main goals of Econometrics.

 Analysis i.e., testing economic theory

 Policy making i.e., obtaining numerical estimates of the
coefficients of economic relationships for policy
simulations.
 Forecasting i.e., using the numerical estimates of the
coefficients in order to forecast the future values of
economic magnitudes.
 Use current and past economic data to predict future values of
variables such as inflation, GDP, stock prices, etc.
 Review questions

• How would you define econometrics?

• How does it differ from mathematical economics and statistics?
• Describe the main steps involved in any econometrics research.
• Differentiate between economic and econometric model.
• What are the goals of econometrics?
UNIT 2: CORRELATION
THEORY

Basic concepts of Correlation

Types of correlation

Methods and Types of Measuring Correlation

 BASIC CONCEPTS OF CORRELATION

Economic variables have a great tendency of moving together

There is a possibility that the change in one variable is on

average accompanied by the change of the other variable. This
situation is known as correlation.

Correlation may be defined as the degree of linear relationship

existing between two or more variables.

The degree of relationship existing between two variables is

called simple correlation.
 CONT’D…
The degree of relationship connecting three or more variables is
called multiple correlations.

A correlation is also said to be partial if it studies the degree of

relationship between two variables keeping all other variables
connected with these two are constant.

Correlation may be linear, when all points (X, Y) on scatter

diagram seem to cluster near a straight, or nonlinear, when all
points seem to lie near a curve.
 CONT’D…

Correlation is said to be linear if the change in one variable

brings a constant change of the other.

It may be non-linear if the change in one variable brings a

different change in the other.

Correlation may also be positive or negative.

CONT’D…
For example, the correlation between price of a commodity and
its quantity supplied is positive since as price rises, quantity
supplied will be increased and vice versa.

Correlation is said to negative if an increase or a decrease in one

variable is accompanied by a decrease or an increase in the other
in which both are changed with opposite direction.

For example, the correlation between price of a commodity and

its quantity demanded.
 METHODS AND TYPES OF MEASURING CORRELATION

 In correlation analysis there are two important things to be

addressed.

• These are the type of co-variation existed between variables and its
strength.

There are three methods of measuring correlation. These are:

 The Scattered Diagram or Graphic Method
 The Simple Linear Correlation coefficient
 The coefficient of Rank Correlation
 1. THE SCATTERED DIAGRAM OR
GRAPHIC METHOD
The scatter diagram is a rectangular diagram which can help us in visualizing the
relationship between two phenomena.
 It puts the data into X-Y plane by moving from the lowest data set to the
highest data set.
 It is a non-mathematical method of measuring the degree of co-variation
between two variables.
 Scatter plots usually consist of a large body of data.
 CONT’D…
 The closer the data points come together and make a straight line, the higher
the correlation between the two variables, or the stronger the relationship.

 If the data points make a straight line going from the origin out to high x- and
y-values, then the variables are said to have a positive correlation.

 If the line goes from a high-value on the y-axis down to a high-value on the
x-axis, the variables have a negative correlation.
 CONT’D…
 A perfect positive correlation = 1, perfect negative correlation = -1.

 no correlation = 0.

 The closer the number is to 1 or -1, the stronger the correlation, or the
stronger the relationship between the variables.

 The closer the number is to 0, the weaker the correlation.

 Two variables may have a positive correlation, negative correlation, or they

may be uncorrelated.
 CONT’D…
 Two variables are said to be positively correlated if they tend to
change together in the same direction, that is, if they tend to
increase or decrease together.

 Two variables are said to be negatively correlated if they tend to

change in the opposite direction
 CONT’D…

 When X increases Y decreases, and vice versa.

 For example, saving and household size are negatively correlated. When HH
size increases, saving decreases and, vice versa.

 The scatter diagram indicates the strength of the relationship between the two
variables.
Provide u’r own example for +
vely and – vely correlated
variables. Or find independent
variables that are + vely and – vely
 2. SIMPLE CORRELATION

For a precise quantitative measurement of the degree of correlation between Y

and X we use a parameter which is called the correlation coefficient.

For example if we measure the correlation between X and Y the population

correlation coefficient is represented by xy and its sample estimate by rxy.

The simple correlation coefficient is used to measure relationships which are

simple and linear only.
 CONT’D…
It cannot help us in measuring non-linear as well as multiple correlations.

Sample correlation coefficient is defined by the formula

n X i Yi   X i Y i

r
n X i  ( X i ) 2 n Yi  ( Yi ) 2
2 2

x i yi
rxy 
x y
2 2
i i

Where, xi  X i  X and y i  Yi - Y
 CONT’D…
Example 2.1: The following table shows the quantity supplied for
a commodity with the corresponding price values.

Q . Determine the type of correlation that exists

between these two variables.

Table 1: Data for computation of correlation coefficient

 CONT’D…
Time period(in Quantity supplied Yi Price Xi (in Birr)
days) (in tons)
1 10 2
2 20 4
3 50 6
4 40 8
5 50 10
6 60 12
7 80 14
8 90 16
9 90 18
10 120 20
 CONT’D…
Y X xi  X i  X y i  Yi  Y x2 y2 xiyi XY X2 Y2
10 2 -9 -51 81 2601 459 20 4 100
20 4 -7 -41 49 1681 287 80 16 400
50 6 -5 -11 25 121 55 300 36 2500
40 8 -3 -21 9 441 63 320 64 1600
50 10 -1 -11 1 121 11 500 100 2500
60 12 1 -1 1 1 -1 720 144 3600
80 14 3 19 9 361 57 1120 196 6400
90 16 5 29 25 841 145 1440 256 8100
90 18 7 29 49 841 203 1620 324 8100
120 20 9 59 81 3481 531 2400 400 14400
Sum=610 110 0 0 330 10490 1810 8520 1540 47700
Mean=61 11
 CONT’D…
n XY   X  Y
10(8520)  (110 )(610)
r   0.975
10(1540)  (110 )(110 ) 10(47700)  (610)(610)
n  X 2  ( X ) 2 n  Y 2  ( Y ) 2

Or using the deviation form (Equation 2.2), the correlation coefficient can be
computed as:
1810
r  0.975
330 10490
 CONT’D…
 There is a strong positive correlation between the quantity supplied and the
price of the commodity under consideration.

 NB: If the correlation coefficient is zero, it indicates that there is no linear

relationship between the two variables.

 If the two variables are independent, the value of correlation coefficient is zero
but zero correlation coefficient does not show us that the two variables are
independent.
 PROPERTIES OF SIMPLE CORRELATION COEFFICIENT

The value of correlation coefficient always ranges between -1 and +1.

The correlation coefficient is symmetric. That means, the correlation coefficient

of X on Y, is the correlation coefficient of Y on X.

The correlation coefficient is independent of change of origin and change of

scale.
 CONT’D …

If X and Y variables are independent, the correlation coefficient is zero. But
the converse is not true.

The correlation coefficient has the same sign with that of regression
coefficients.

The correlation coefficient is the geometric mean of two regression

coefficients.
 THE MAJOR LIMITATIONS OF THE
METHOD ARE:
The value of the coefficient is unduly affected by the extreme
values

The coefficient requires the quantitative measurement of both

variables.

The correlation coefficient always assumes linear relationship

regardless of the fact whether the assumption is true or not.

The coefficient is misinterpreted.

 3. THE RANK CORRELATION
COEFFICIENT

In many cases the variables may be qualitative (or binary variables) and
hence cannot be measured numerically.

For example, profession, education, preferences for particular brands, are

such categorical variables.

For such cases it is possible to use another statistic, the rank correlation
coefficient (or spearman’s correlation coefficient.).

We rank the observations in a specific sequence for example in order of size,
importance, etc., using the numbers 1, 2, 3, …, n.
 CONT’D…
In other words, we assign ranks to the data and measure relationship between
their ranks instead of their actual numerical values.

Hence, the name of the statistic is given as rank correlation coefficient.

If two variables X and Y are ranked in such way that the values are ranked in
ascending or descending order, the rank correlation coefficient may be
computed by the formula
 CONT’D…

Where,
D = difference between ranks of corresponding pairs of X and Y
n = number of observations.

 The values that r may assume range from + 1 to – 1.

 Two points are of interest when applying the rank correlation coefficient.
 It does not matter whether we rank the observations in ascending or
descending order.
 However, we must use the same rule of ranking for both variables.
 CONT’D…EXAMPLE

Brands of A B C D E F G H I J K L Total
soap

Person I 9 10 4 1 8 11 3 2 5 7 12 6

Person II 7 8 3 1 10 12 2 6 5 4 11 9

Di 2 2 1 0 -2 -1 1 -4 0 3 1 -3

Di2 4 4 1 0 4 1 1 16 0 9 1 9 50
 CONT’D…
The rank correlation coefficient

This figure, 0.827, shows a marked similarity of preferences of the two

persons for the various brands of soap.
4. PARTIAL CORRELATION
COEFFICIENTS
A partial correlation coefficient measures the relationship
between any two variables, when all other variables connected
with those two are kept constant.
For example, let us assume that we want to measure the
correlation between the number of hot drinks (X1) consumed in
a summer resort and the number of tourists (X2) coming to that
resort.
In order to measure the true correlation between X1 and X2, we
must find some way of accounting for changes in X3.
 CONT’D…
This is achieved with the partial correlation coefficient between
X1 and X2, when X3 is kept constant.

The partial correlation coefficient is determined in terms of the

simple correlation coefficients among the various variables
involved in a multiple relationship.

In our example there are three simple correlation coefficients

 CONT’D…
r12 = correlation coefficient between X1 and X2
r13 = correlation coefficient between X1 and X3
r23 = correlation coefficient between X2 and X3

The partial correlation coefficient between X1 and X2, keeping the effect of X3
constant is given by:
EXAMPLE
The following table gives data on the yield of corn per acre(Y),
the amount of fertilizer used(X1) and the amount of insecticide
used (X2). Compute the partial correlation coefficient between
the yield of corn and the fertilizer used keeping the effect of
insecticide constant.
Year 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980

Y 40 44 46 48 52 58 60 68 74 80

X1 6 10 12 14 16 18 22 24 26 32

X2 4 4 5 7 9 12 14 20 21 24
 ANSWER
Year Y X1 X2 y x1 x2 x1y x2y x1x2 x12 x22 y2
1971 40 6 4 -17 -12 -8 204 136 96 144 64 289
1972 44 10 4 -13 -8 -8 104 104 64 64 64 169
1973 46 12 5 -11 -6 -7 66 77 42 36 49 121
1974 48 14 7 -9 -4 -5 36 45 20 16 25 81
1975 52 16 9 -5 -2 -3 10 15 6 4 9 25
1976 58 18 12 1 0 0 0 0 0 0 0 1
1977 60 22 14 3 4 2 12 6 8 16 4 9
1978 68 24 20 11 6 8 66 88 48 36 64 121
1979 74 26 21 17 8 9 136 153 72 64 81 289
1980 80 32 24 23 14 12 322 276 168 196 144 529
Sum 570 180 120 0 0 0 956 900 524 576 504 1634
Mean 57 18 12
 ANSWER
ryx1 = 0.9854

ryx = 0.9917
2

rx1x = 0.9725
2

Then
Any Q uestion …

Q
UNIT 3: SIMPLE LINEAR
REGRESSION MODELS

Basic Concepts and Assumptions

Least Squares Criteria

Normal Equations of OLS

Coefficient of Correlation and Determination

Properties of OLS Estimators

 Hypothesis Testing
BASIC CONCEPTS AND
ASSUMPTIONS

There are two major ways of estimating regression functions.

ordinary least square (OLS) method and
maximum likelihood (MLH) method.
The ordinary least square method is the easiest and the most commonly used
method as opposed to the maximum likelihood (MLH) method which is limited
by its assumptions.
MLH method is valid only for large sample as opposed to the OLS method
which can be applied to smaller samples.
 CONT’D…
Classical Assumptions
The error terms ‘Ui’ are randomly distributed or the disturbance terms are not
correlated. This means that there is no systematic variation or relation among the
value of the error terms (Ui and Uj); Where i = 1, 2, 3, …….., j = 1, 2, 3, …….
and .
Cov (Ui , Uj) = 0 for population
Cov (ei, ej) = 0 for sample .
The disturbance terms ‘Ui’ have zero mean. This implies the sum of the
individual disturbance terms is zero.
 CONT’D…
This results in an upward or down ward shift in the
regression function.
 CONT’D…
The disturbance terms have constant variance in each period
(homoscedasticity).
if the variance of the error terms varies as sample size changes or as the value of
explanatory variables changes, then this leads to heteroscedasticity problem.

Explanatory variables ‘Xi’ and disturbance terms ‘Ui’ are uncorrelated or

independent.
 CONT’D…
All the co-variances of the successive values of the error term are equal to zero.

 The value in which the error term assumed in one period does not depend on
the value in which it assumed in any other period. non-autocorrelation or non-
serial correlation.

 The explanatory variable Xi is fixed in repeated samples (the explanatory

variables are non-random).

 Each value of Xi does not vary for instance owing to change in sample
size.
 CONT’D…
 Linearity of the model in parameters. The simple linear regression requires
linearity in parameters; but not necessarily linearity in variables(what is
important is transforming the data as required).

 Normality assumption-The disturbance term Ui is assumed to have a normal

distribution with zero mean and a constant variance.

 Explanatory variables should not be perfectly, linearly and/or highly

correlated.
 CONT’D…

The relationship between variables (or the model) is correctly specified.

The explanatory variables do not have identical value. This assumption is very
important for improving the precision of estimators.
 OLS METHOD OF ESTIMATION

Estimating a linear regression function using the Ordinary Least

Square (OLS) method is simply about calculating the parameters
of the regression function for which the sum of square of the error
terms is minimized.
Suppose we want to estimate the following equation
Sample regression:
 CONT’D…

RSS= Residual Sum of Squares.

Note that the equation is a composite function and we should

apply a chain rule in finding the partial derivatives with respect
to the parameter estimates.
CONT’D…

That is, the partial derivative with respect to ̂ 0

CONT’D…

n XY  ( X )(  Y )
ˆ1 
Or
n X 2  ( X ) 2

_ _
  XY  n Y X and we have ˆ 0  Y  ˆ1 X from above
1  _
 X i2  n X 2
 CONT’D…

Example: Given the following sample data of three pairs of ‘Y’

(dependent variable) and ‘X’ (independent variable).

Yi Xi
10 30
20 50
30 60
CONT’D…

A. find a simple linear regression function;

Y = f(X)

B. Interpret your result.

C. Predict the value of Y when X is 45.

 CONT’D…
Solution
a. To fit the regression equation we do the following computations.
Yi Xi Yi Xi Xi2
10 30 300 900
20 50 1000 2500
30 60 1800 3600
Sum 60 140 3100 7000
Mean Y = 20 140
X=
3

n XY  ( X )(  Y )
3(3100)  (140)( 60)
ˆ1    0.64
3(7000)  (140) 2
n X 2  ( X ) 2

ˆ0  Y  ˆ1 X  20  0.64(140 / 3)  10

Thus the fitted regression function is given by: Yˆi   10  0.64 X i

 CONT’D…
B. Interpretation, the value of the intercept term, -10, implies that the
value of the dependent variable ‘Y’ is – 10 when the value of the
explanatory variable is zero.

The value of the slope coefficient is a measure of the marginal

change in the dependent variable ‘Y’ when the value of the
explanatory variable increases by one.

 For instance, in this model, the value of ‘Y’ increases on average by

0.64 units when ‘X’ increases by one.

C. Y= -10+(0.64) (45)=18.8
 CONT’D…
 That means when X assumes a value of 45, the value of Y on
average is expected to be 18.8.

 The regression coefficients can also be obtained by simple

formulae by taking the deviations between the original values and
their means.

Then, the coefficients can be represented by alternative formula:

 CONT’D…
Example 2.5: Find the regression equation for the data under Example 2.4, using the
shortcut formula. To solve this problem we proceed as follows.
Yi Xi y x xy x2 y2
10 30 -10 -16.67 166.67 277.78 100
20 50 0 3.33 0.00 11.11 0
30 60 10 13.33 133.33 177.78 100
Sum 60 140 0 0 300.00 466.67 200
Mean 20 46.66667

Then

x i yi
300 , and ˆ0  Y  ˆ1 X =20-(0.64) (46.67) = -10 with
ˆ1    0.64
466.67
x
2
i

results similar to previous case.

 PROPERTIES OF OLS ESTIMATORS

 There can be several samples of the same size that can be

drawn from the same population.
 For each sample, the parameter estimates have their own
specific numerical values. Therefore, the parameter estimate
have different values for estimating a given true population
parameter.
 Thus, the parameter estimates also have a normal distribution
with their associative mean and variance.
 THE MEAN AND VARIANCE OF THE
PARAMETER ESTIMATES

Formula for mean and variance of the respective parameter estimates and the
error term are given below:
1. The mean of

2. The variance of

3. The mean of
4. The variance of
5. The estimated value of the variance of the error term
STANDARD ERROR OF THE
PARAMETERS
HYPOTHESIS TESTING

After estimation of the parameters there are hypothesis testing.

We have to know that to what extent our estimates are reliable
enough and/or acceptable for further purpose.

That means, we have to evaluate the degree of

representativeness of the estimate to the true population
parameter.

Simply a model must be tested for its significance before it can

be used for any other purpose.
 CONT’D…
The available test criteria are divided in to three groups
(Theoretical a priori criteria, statistical criteria and
econometric criteria).
1. Priori criteria set by economic theories are in line with the
consistency of coefficients of econometric model to the
economic theory.
2. Statistical criteria(first order tests), are set by statistical
theory and refer to evaluate the statistical reliability of the
model.
3. Econometric criteria refer to whether the assumptions of an
econometric model employed in estimating the parameters
are fulfilled or not.
CONT’D…

bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb
bbbbbbbbbbbbbbbbbbbb
 THE COEFFICIENT OF
DETERMINATION (R 2 )

The coefficient of determination is the measure of the amount or

proportion of the total variation of the dependent variable that is
determined or explained by the model or the presence of the
explanatory variable in the model.

The total variation of the dependent variable is measured from

its arithmetic mean.
 CONT’D…

The total variation of the dependent variable is given in the following form;

TSS=ESS + RSS

o which means total sum of square of the dependent variable is split into
explained sum of square and residual sum of square.
 CONT’D…
The coefficient of determination is given by the formula

 (Yˆ  yˆ
2
i  Y )2 i
Explained Variation in Y
R2   
Total Variation in Y
 (Y y
2
i  Y )2 i

2.16
 2 
Since  y i   1  x i y i the coefficient of determination can also be given as

 1  xi y i
R2 
 y i2

Or

 (Y  Yˆi ) 2 e
2
i i
Unexplained Variation in Y
R2  1  1  1
Total Variation inY
 (Y y
2
i Y ) 2
i
 CONT’D…

The higher the coefficient of determination is the better the fit.

Conversely, the smaller the coefficient of determination is the
poorer the fit.
One minus the coefficient of determination is called the
coefficient of non-determination, and it gives the proportion of
the variation in the dependent variable that remained
undetermined or unexplained by the model.
 TESTING THE SIGNIFICANCE OF
REGRESSION COEFFICIENT

Since the sample values of the intercept and the coefficient are
estimates of the true population parameters, we have to test them
for their statistical reliability.

The significance of a model can be seen in terms of the amount of

variation in the dependent variable that it explains and the
significance of the regression coefficients.
 CONT’D…

 There are different tests that are available to test the statistical
reliability of the parameter estimates. The following are the
common ones;

 The standard error test

 The students t-test
 The standard normal test
 1. THE STANDARD ERROR TEST

This test first establishes the two hypotheses (the null and
alternative hypotheses). The two hypotheses are given as follows:
H0: βi=0
H1: βi≠0
The standard error test is outlined as follows:
1. Compute the standard deviations of the parameter estimates
using the formula
 CONT’D…

Compare the standard errors of the estimates with the

regression coefficients and make decision.
A) If the standard error of the estimate is less than half of the
absolute regression coefficients, that the estimate is statistically
significant.
If so, reject the null hypothesis and we can conclude that the
estimate is statistically significant.
 CONT’D…

B) If the standard error of the estimate is greater than half of the

numerical value of the estimate, the parameter estimate is not
statistically reliable.

If so, conclude to accept the null hypothesis and conclude that

the estimate is not statistically significant.
 2. THE STUDENT T-TEST

In conditions where Z-test is not applied, t-test can be used to test

the statistical reliability of the parameter estimates and vice versa.

The test depends on the degrees of freedom that the sample has.
The test procedures of t-test are similar with that of the z-test.
 CONT’D…
The procedures are outlined as follows;
 Set up the hypothesis.
 Determine the level of significance (usually a 5% )level)
 Determine the tabulated value of t from the table with n-k degrees of
freedom, where k is the number of parameters estimated.
 Determine the calculated value of t.

The test rule or decision is given as follows:

Reject H0 if

| t cal | t / 2,n  k
ˆi
tcal 
se( ˆi )
 3. THE STANDARD NORMAL TEST

This test is based on the normal distribution.

The test is applicable if:
The standard deviation of the population is known irrespective of the sample
size
The standard deviation of the population is unknown provided that the sample
size is sufficiently large (n>30).
 CONT’D…
The standard normal test or Z-test is outlined as follows;
 Test the null hypothesis against the alternative hypothesis
 Determine the level of significant (it is common in applied
econometrics to use 5% level of significance).
 Determine the theoretical or tabulated value of Z from the
table.
 Make decision.
 CONT’D…
The decision of statistical hypothesis testing consists of two
decisions; either accepting the null hypothesis or rejecting it.
If , accept the null hypothesis while if , reject the null hypothesis.
It is true that most of the times the null and alternative hypotheses
are mutually exclusive.
Accepting the null hypothesis means that rejecting the alternative
hypothesis and rejecting the null hypothesis means accepting the
alternative hypothesis.
 CONT’D…
Example: If the regression has a value of =29.48 and the standard error is 36.
Test the hypothesis that the value of at 5% level of significance using standard
normal test.
Solution: We have to follow the procedures of the test.
After setting up the hypotheses to be tested, the next step is to determine the
level of significance in which the test is carried out. In the above example the
significance level is given as 5%.
The third step is to find the theoretical value of Z at specified level of
significance. From the standard normal table we can get that .
The fourth step in hypothesis testing is computing the observed or calculated
value of the standard normal distribution using the following formula.
. Since the calculated value of the test statistic is less than the tabulated value,
the decision is to accept the null hypothesis and conclude that the value of the
parameter is 25.
 CONFIDENCE INTERVAL
ESTIMATION OF THE
REGRESSION COEFFICIENTS
rejecting the null hypothesis does not mean that the parameter
estimates are correct estimates of the true population parameters.
It means that the estimate comes from the sample drawn from the
population whose population parameter is significantly different
from zero.

In order to define the range within which the true parameter
lies, we must construct a confidence interval for the parameter.
we can construct 100(1- ) % confidence intervals for the
sample regression coefficients.
 CONT’D…
NB. The standard error of a given coefficient is the positive square root of
the variance of the coefficient.
X
2
i

 Variance of the intercept is given by var( ˆ0 )   u

2

n xi
2

1
var(ˆ1 )   u
2
 Variance of the slope is given by
x
2
i

e
2
i
Where u2 
nk

is the estimate of the variance of the random term and k is the number of
parameters to be estimated in the model.
 CONT’D…
The standard errors are the positive square root of the variances and the 100 (1-
) % confidence interval for the slope is given by:

   
1  t (n  k )( se( 1 ))  1  1  t (n  k )( se( 1 ))
2 2

1  ˆ1  t / 2,nk ( se( ˆ1 ))

 0  ˆ0  t / 2,nk ( se( ˆ0 ))

 CONT’D…
Example 2.6: The following table gives the quantity supplied (Y in tons) and
its price (X pound per ton) for a commodity over a period of twelve years.

Y 69 76 52 56 57 77 58 55 67 53 72 64

X 9 12 6 10 9 10 7 8 12 6 11 8
 SOLUTION

Data for computation of different parameters

Time Y X XY X2 Y2 x y xy x2 y2 Yˆ ei e i2
1 69 9 621 81 4761 0 6 0 0 36 63.00 6.00 36.00
2 76 12 912 144 5776 3 13 39 9 169 72.75 3.25 10.56
3 52 6 312 36 2704 -3 -11 33 9 121 53.25 -1.25 1.56
4 56 10 560 100 3136 1 -7 -7 1 49 66.25 -10.25 105.06
5 57 9 513 81 3249 0 -6 0 0 36 63.00 -6.00 36.00
6 77 10 770 100 5929 1 14 14 1 196 66.25 10.75 115.56
7 58 7 406 49 3364 -2 -5 10 4 25 56.50 1.50 2.25
8 55 8 440 64 3025 -1 -8 8 1 64 59.75 -4.75 22.56
9 67 12 804 144 4489 3 4 12 9 16 72.75 -5.75 33.06
10 53 6 318 36 2809 -3 -10 30 9 100 53.25 -0.25 0.06
11 72 11 792 121 5184 2 9 18 4 81 69.50 2.50 6.25
12 64 8 512 64 4096 -1 1 -1 1 1 59.75 4.25 18.06
Sum 756 108 6960 1020 48522 0 0 156 48 894 756.00 0.00 387.00
 CONT’D…
1. Estimate the Coefficient of determination (R2), Coefficient of non-
determination (1- R2) and interpret the result.
2. Find the variance and standard error of the intercept
3. Find the variance and standard error of the slope
4. Run significance test of regression coefficients using the following test
methods
The standard error test
The students t-test

5. Fit the linear regression equation and determine the 95% confidence interval
for the slope.
 SOLUTION
1. Estimate the Coefficient of determination (R2)

e
2
i
387
R  1
2
 1  1  0.43  0.57
894
y
2
i

This result shows that 57% of the variation in the quantity supplied of the
commodity under consideration is explained by the variation in the price of the
commodity; and the rest 43% remain unexplained by the price of the
commodity
 CONT’D…
4. Run significance test of regression coefficients using fitted regression line for
the data given:

Yi  33.75  3.25 X i
(8.3) (0.9)

A. Standard Error test

Since the standard error is less than half of the numerical value of the slope,
we have to reject the null hypothesis and conclude that is statistically
significant.
 CONT’D…
B. The students t-test
ˆi 3.25
t cal    3.62
ˆ
se(  i ) 0.8979

Further tabulated value for t is 2.228.

When we compare these two values, the calculated t is greater than the
tabulated value. Hence, we reject the null hypothesis.
Rejecting the null hypothesis means, concluding that the price of the
commodity is significant in determining the quantity supplied for the
commodity.
 CONT’D…
Generally, a two tail test of a null hypothesis at 5% level of significance can
be reduced to the following two t-rules.

1. If tcal is greater than 2 or less than -2, we reject the null hypothesis
2. If tcal is less than 2 or greater than -2, accept the null hypothesis.
 CONT’D…
5. To estimate confidence interval we need standard error which is determined
as follows:
e
2
i
387 387
u2     38.7
nk 12  2 10

1 1
var( ˆ1 )   u
2
 38.7( )  0.80625
48
x 2

se( ˆ1 )  var( ˆ1 )  0.80625  0.8979

ˆ1  3.25  (2.228)(0.8979)  3.25  2  3.25  2, 3.25  2  1.25, 5.25

 PROPERTIES OF OLS ESTIMATORS

The ideal or optimum properties that the OLS estimates possess may
be summarized by well known theorem known as the Gauss-Markov
Theorem.

Statement of the theorem: “Given the assumptions of the classical

linear regression model, the OLS estimators, in the class of linear and
unbiased estimators, have the minimum variance, i.e. the OLS
estimators are BLUE.
 CONT’D…

The least squares estimators are linear, unbiased and have minimum variance

(i.e. are best of all linear unbiased estimators).

An estimator is called BLUE (Best, Linear, Unbiased Estimator) if it is:

Linear: a linear function of the random variable, such as, the dependent
variable Y.

Unbiased: its average or expected value is equal to the true population

parameter.
 CONT’D…

Minimum variance: It has a minimum variance in the class of linear and

unbiased estimators.

NB: An unbiased estimator with the least variance is known as an efficient

estimator.

According to the Gauss-Markov theorem, the OLS estimators possess all the
BLUE properties.
Any Q uestion …

Q
QUIZ

Given the following sample data of three pairs of ‘Y’

(dependent variable) and ‘X’ (independent variable).

Yi Xi
10 30
20 50
30 60
 QUIZ…

1. Fit the regression equation

2. Estimate the Coefficient of determination (R2), Coefficient of non-determination
(1- R2) and interpret the result.
3. Find the variance and standard error of the intercept
4. Find the variance and standard error of the slope
5. Run significance test of regression coefficients using the following test methods
The standard error test
The students t-test

5. Fit the linear regression equation and determine the 95% confidence interval for the
slope.
UNIT 4: MULTIPLE REGRESSION
ANALYSIS

 Multiple Regression Models

 Notations and Assumptions
 Estimation of Partial regression coefficients
 Analysis of Variance
 Hypothesis Testing
 Dummy Variable Regression Analysis
 MULTIPLE REGRESSION MODEL
The multiple linear regression (population regression function) in which we
have one dependent variable Y, and k explanatory variables
X 1 , (X 2 ,...... X k ), is
given by:

Yi   0  1 X 1   2 X 2  ....   k X k  u i

Where,  0  the intercept = value of Y when all X’s are zero

 i = are partial slope coefficients
u i = the random term
CONT’D…

But for the sake of simplicity possible regression model with

two-variables regression is presented as:

Yi   0   1 X 1   2 X 2  u i
 NOTATIONS AND ASSUMPTIONS

Randomness of ui - the variable u is a real random variable.

Zero mean of ui - the random variable has a zero mean for each value
Homoscedasticity of the random term - the random term has constant
variance. In other words, the variance is the same for all the explanatory
values.
Normality - the values of each are normally distributed

E (u i )  0

E (u i2 )   u2 Cons tan t

u i  N (0,  u2 )
 ESTIMATION OF PARTIAL
REGRESSION COEFFICIENTS

the sample regression function will look like the following.

^ ^ ^ ^
Yi   0  1 X 1   2 X 2
^
ei  Yi  Y i

^ ^ ^
 0 ,  1 and  2 in such a way that i
e 2
Minimum
n
  (Y  ˆ0  ˆ1 X 1  ˆ 2 X 2 ) 2
 e 2
i
 i 1
0 3.5
ˆ0
^
 0
n
  (Y  ˆ0  ˆ1 X 1  ˆ 2 X 2 ) 2
 e 2
i
 i 1
0 3.6
ˆ1
^
 1
n
  (Y  ˆ0  ˆ1 X 1  ˆ 2 X 2 ) 2
 e 2
i
 i 1
0 3.7
ˆ 2
^
 2
 ESTIMATION OF PARTIAL
REGRESSION COEFFICIENTS
Solving equations (3.5), (3.6) and (3.7) simultaneously, we obtain the system of
normal equations given as follows:

^ ^ ^
Y  n     X    X
i 0 1 1i 2 2i 3 .8
^ ^ ^
X Y   X  X  X X
1i i 0 1i 1 1
2
2 1i 2i 3 .9
^ ^ ^
X Y   X  X X b X
2i i 0 2i 2 1i 2i 2
2
2i 3.10
 CONT’D…
Then Letting

x1i  X 1i  X 1 3.11

x 2i  X 2i  X 2 3.12
yi  Yi  Y 3.13

^  x y  x   x y  x x 
1 1
2
2 2 1 2
1  3.14
 x  x   x x 
2
1
2
2 1 2
2

^  x y  x   x y  x x 
2
2
1 1 1 2
2  3.15
 x  x   x x 
2
1
2
2 1 2
2

^   
 0  Y  ˆ1 X 1  ˆ 2 X 2 3.16
 VARIANCE AND STANDARD ERRORS
OF OLS ESTIMATORS

 Estimating the numerical values of the parameters is not enough

in econometrics if the data are coming from the samples.
 The standard errors derived are important for two main
purposes:
 To establish confidence intervals for the parameters
and
 To test statistical hypotheses.

 2
 2   
 X 1  x 2  X 2  x1  2 X 1 X 2  x1 x 2 
2 2
^2 1
 
Var   0    ui   
^

n 
 
  x1  x 2  ( x1 x 2 )
2 2 2

 
CONT’D…
^ ^
SE (  0 )  Var (  0 )

 ^  
Var  1    u 
2  x 22  ^ ^
2 
    x1  x 2  ( x1 x 2 ) 
2 2
SE ( 1 )  Var ( 1 )

^ ^ 2 
Var (  2 )   u 
 x12  ^ ^
2 
  x1  x 2  ( x1 x 2 ) 
2 2 SE (  2 )  Var (  2 )


^
2

 e 2
i
u
n3
 COEFFICIENT OF MULTIPLE
DETERMINATION
 Is the measure of the proportion of the variation in the dependent variable
that is explained jointly by the independent variables in the model.

R2 
y 2

3.24
y 2

 However, every time we insert additional explanatory variable in the model,

the increases irrespective of the improvement in the goodness-of- fit of the
model

^ ^
 1  x1 y   2  x 2 y
R2  3.25
 y2
 CONT’D…
 That means, highest value may not imply that the model is good.

(n  1)
2
Rady  1  (1  R 2 ) 3.26
(n  k )

K is number of parameters
2
 In multiple linear regression, therefore, we better interpret the adjusted R
2
than the ordinary or the unadjusted R .
 CONFIDENCE INTERVAL
ESTIMATION
Interpretation of the confidence interval: Values of the parameter lying in the
interval are plausible with 100(1-  )% confidence.

ˆi  t / 2,n  k se( ˆi )

 HYPOTHESIS TESTING

 Testing hypothesis about an individual partial regression coefficient;

 Testing the overall significance of the estimated multiple regression

model;

 Testing if two or more coefficients are equal to one another;

Testing the stability of the estimated regression model over time/cross-

sectional units
 TESTING INDIVIDUAL REGRESSION
COEFFICIENTS

using the standard error test or the t-test

H 0 : ˆ1  0 H 0 : ˆ 2  0 H 0 : ˆ K  0
H : ˆ  0 H : ˆ  0
1 2
H 1 : ˆ K  0
1 1

A. Standard Error Test: decision rule is based on the relationship between the
numerical value of the parameter and the standard error of the same parameter.
1
S ( ˆ i )  ˆ i
If , 2 we reject the null hypothesis, i.e. the estimate is statistically
significant.

Generalisation: The smaller the standard error, the stronger is the evidence that
the estimates are statistically significant.
 CONT’D…

B. t-test – the more appropriate and formal way to test the hypothesis.
compute the t-ratios and compare them with the tabulated t-values and make our
decision.

Decision Rule: accept H0 if tcal < ttab

 Rejecting H0 means, the coefficient being tested is significantly different from 0.

Accepting H0, on the other hand, means we don’t have sufficient evidence to
conclude that the coefficient is different from 0.
TESTING THE OVERALL SIGNIFICANCE OF REGRESSION MODEL

Hypotheses of such type are often called joint hypotheses.

Testing the overall significance of the model means testing the
null hypothesis that none of the explanatory variables in the
model significantly determine the changes in the dependent
variable.

H 0 : 1   2  0
H 1 :  i  0, at least for one i.
 CONT’D…

The test statistic for this test is given by:

 yˆ 2

Fcal  k 1

e 2

nk

Where, k is the number of parameters in the model.

 CONT’D…
Overall significance test of a model are summarized in the
analysis of variance (ANOVA) table as follows:

Source of variation Sum of squares Degrees of Mean sum of Fcal

freedom squares
Regression ^2 k 1 ^
MSE
SSE   y MSE 
y 2
F
k 1 MSR
Residual SSR   e 2 nk e 2

MSR 
nk
Total SST   y 2 n 1
 CONT’D…

These three sums of squares:

^
SSE   y   (Yˆi  Y ) 2  Explained Sum of Squares
2

SSR   ey i2   (Yi  Yˆ ) 2  Unexplained Sum of Squares

SST   y 2   (Yi  Y ) 2  Total Sum of Squares

SST  SSE  SSR

The test rule: reject H0 if Fcal ≥ Ftab

RELATIONSHIP BETWEEN F AND
R2
 yˆ 2

R 2

y 2  e 2
 (1  R 2
)  y 2

 yˆ 2
 R 2
y 2
Fcal 
 y2
k 1
 e2
R2  1
 e 2
nk
y 2

R2  y2 R2  y2
 e2  1 R 2 Fcal   .
(n  k )
k 1 (1  R 2 ) y 2
 y2 (1  R 2 ) y 2
k 1

nk
(n  k ) R 2
Fcal  .
k  1 (1  R 2 )
 EXAMPLE

The following table shows a particular country’s imports (Y),

the level of Gross National Product(X1) measured in arbitrary

units, and the price index of imported goods (X2), over 12

years period.
CONT’D…

Year 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971

Y 57 43 73 37 64 48 56 50 39 43 69 60

X1 220 215 250 241 305 258 354 321 370 375 385 385

X2 125 147 118 160 128 149 145 150 140 115 155 152
 CONT…
a. Estimate the coefficients of the economic relationship and fit the model.
Year Y X1 X2 x1 x2 y X12 x 22 x 1y x 2y x 1x 2 y2

1960 57 220 125 -86.5833 -15.3333 3.75 7496.668 235.1101 -324.687 -57.4999 1327.608 14.0625

1961 43 215 147 -91.5833 6.6667 -10.25 8387.501 44.44489 938.7288 -68.3337 -610.558 105.0625

1962 73 250 118 -56.5833 -22.3333 19.75 3201.67 498.7763 -1117.52 -441.083 1263.692 390.0625

1963 37 241 160 -65.5833 19.6667 -16.25 4301.169 386.7791 1065.729 -319.584 -1289.81 264.0625

1964 64 305 128 -1.5833 -12.3333 10.75 2.506839 152.1103 -17.0205 -132.583 19.52731 115.5625

1965 48 258 149 -48.5833 8.6667 -5.25 2360.337 75.11169 255.0623 -45.5002 -421.057 27.5625

1966 56 354 145 47.4167 4.6667 2.75 2248.343 21.77809 130.3959 12.83343 221.2795 7.5625

1967 50 321 150 14.4167 9.6667 -3.25 207.8412 93.44509 -46.8543 -31.4168 139.3619 10.5625

1968 39 370 140 63.4167 -0.3333 -14.25 4021.678 0.111089 -903.688 4.749525 -21.1368 203.0625

1969 43 375 115 68.4167 -25.3333 -10.25 4680.845 641.7761 -701.271 259.6663 -1733.22 105.0625

1970 69 385 155 78.4167 14.6667 15.75 6149.179 215.1121 1235.063 231.0005 1150.114 248.0625

1971 60 385 152 78.4167 11.6667 6.75 6149.179 136.1119 529.3127 78.75022 914.8641 45.5625

Sum 639 3679 1684 0.0004 0.0004 0 49206.92 2500.667 1043.25 -509 960.6667 1536.25

Mean 53.25 306.5833 140.3333 0 0 0

 CONT…

a. Estimate the parameters and fit the model.

b. Compute the variance and standard errors of the slopes.

c. Calculate and interpret the coefficient of determination (the adjusted

one).

d. Test the overall significance of the model.

CONT’D…

 Y  639 X 1  3679 X 2  1684 n  12

Y 639
Y    53.25
n 12

X 1
3679
X1    306.5833
n 12

X 2
1684
X2    140.3333
n 12
The summary results in deviation forms are then given by:

x x
2 2
1  49206.92 2  2500.667

 x y  1043.25
1 x 2 y  509

x x1 2  960.6667 y 2
 1536.25
CONT’D…
 CONT’D…
The fitted model is then written as: Yˆi = 75.40512 + 0.025365X1 - 0.21329X2
a) Compute the variance and standard errors of the slopes.
First, you need to compute the estimate of the variance of the random term as follows


^
2

e 2
i

1401.223 1401.223
  155.69143
u
n3 12  3 9
^
Variance of  1

 ^  2

Var   1    u 
 x 22 
  155.69143(
2500.667
)  0.003188
  x1  x 2  ( x1 x 2 ) 
2 2 2
  12212724
^
Standard error of  1
^ ^
SE (  1 )  Var (  1 )  0.003188  0.056462
^
Variance of  2
^ 
Var (  2 )   u 
^ 2
 x12 
 155.69143(
49206.92
)  0.0627
2 
  x1  x 2  ( x1 x 2 ) 
2 2
122127241
^
Standard error of  2
^ ^
SE (  2 )  Var (  2 )  0.0627  0.25046
 CONT’D…
Similarly, the standard error of the intercept is found to be 37.98177. The detail is left for you as
an exercise.
a) Calculate and interpret the coefficient of determination.
We can use the following summary results to obtain the R 2.

 yˆ 2
 135.0262

e 2
 1401.223

y 2
 1536.25 (The sum of the above two). Then,

^ ^
 1  x1 y   2 x 2 y (0.025365)(1043.25)  (-0.21329)(-509)
R 
2
  0.087894
y 2
1356.25

e 2

1401.223
or R 2  1   1  0.087894
1356.25
y 2

b) Compute the adjusted R2.

(n  1 12 - 1
2
R ady  1  (1  R 2 )  1 - (1 - 0.087894)  0.114796
n  k) 12 - 3
 CONT’D…
a) Test the significance of X1 and X2 in determining the changes in Y using t-test.
The hypotheses are summarized in the following table.
Coefficient Hypothesis Estimate Std. error Calculated t Conclusion

1 H0: 1=0 0.025365 0.056462 0.025365 We do not

t cal   0.449249
H1: 10 0.056462 reject H0 since
tcal<ttab
2 H0: 2=0 -0.21329 0.25046  0.21329 We do not
t cal   0.85159
H1: 20  0.21329 reject H0 since
tcal<ttab

The critical value (t 0.05, 9) to be used here is 2.262. Like the standard error test, the t- test revealed
that both X1 and X2 are insignificant to determine the change in Y since the calculated t values
are both less than the critical value.
 CONT’D…
a) Test the overall significance of the model. (Hint: use  = 0.05)
This involves testing whether at least one of the two variables X 1 and X2 determine the changes
in Y. The hypothesis to be tested is given by:
H 0 : 1   2  0
H 1 :  i  0, at least for one i.
The ANOVA table for the test is give as follows:
Source of Sum of Squares Degrees of Mean Sum of Squares Fcal
variation freedom
Regression ^ 2 ^ MSR
SSR   y  135.0262
k  1 =3-1=2 MSR 
 y2 
135.0262

F 
MSE
k 1 2  0.433634
67.51309
Residual SSE   e 2  1401.223 n  k =12- e 2
1401.223
MSE   
3=9 nk 9
155.614
Total SST   y  1536.25
2 n  1 =12-
1=11

The tabulated F value (critical value) is F(2, 11) = 3.98

 CONT’D…

Or…
(n  k ) R2 (12 - 3) 0.087894
Fcal  .   0.433632
k  1 (1  R )
2
3 - 1 1  0.087894

The calculated F value (0.4336) is less than the tabulated value (3.98). Hence,
we accept the null hypothesis and conclude that there is no significant
contribution of the variables X1 and X2 to the changes in Y.
Any Q uestion …

Q
 EXTENSIONS OF REGRESSION MODELS

The relationship between Y and X can be non-linear rather than linear.

The choice of a functional form for an equation is a vital part of the
specification of that equation.
The choice of a functional form almost always should be based on an
examination of the underlying economic theory.
the one that comes closest to that underlying theory should be chosen for
the equation.
 CONT’D…
Some Commonly Used Functional Forms
The Linear Form: It is based on the assumption that the slope of the
relationship between the independent variable and the dependent variable is
constant.

In this case elasticity is not

Y
constant.
 i
X

Economic theory frequently predicts only the sign of a relationship and not its
functional form.N Y / Y Y X X

Y ,X     i
X / X X Y Y
 CONT’D…
Log-linear, double Log or constant elasticity model
The most common functional form that is non-linear in the variable (but still
linear in the coefficients) is the log-linear form.
 A log-linear form is often used, because the elasticities and not the slopes are
constant i.e.,  =  Constant.
CONT’D…
dem and
Output

 
Yi   0X i
1

p r ic e
Input


Yi   0 X i i eU i gd(log f)

ln Yi  ln  0   i ln X i  U i ln Yi  ln  0  1 ln X i

log f price
 CONT’D…
The model is also called a constant elasticity model because the coefficient
of elasticity between Y and X (1) remains constant.

Y X d ln Y
   1
X Y d ln X
 CONT’D…
Semi-log Form

Yi   0   1 ln X 1i  U i ln Yi   0   1 X 1i  U i

1<0
Y=0+1Xi

1>0

x
 CONT’D…
Polynomial Form

Y   0   1 X 1i   2 X 1i   3 X 2i  U i
2

Y
 1  2 2 X 1
X 1

Y Y
 3 Y
X 2

A)
B)

X
Xi
Impact of age on earnings a typical cost curve
 CONT…
Reciprocal Transformation (Inverse Functional Forms)

1
Yi   0   1 ( )   2 X 2i  U i
X 1i
1
Yi   0   1 ( )   2 X 2i  U i
X 1i

1 0  0
Y  0 
X1i 1  0

0

1 0  0
Y  0 
X1i 1  0
DUMMY VARIABLE REGRESSION
ANALYSIS
Dummy variables are discrete variables taking a value of ‘0’ or ‘1’. They are
often called ‘on’ ‘off’ variables, being ‘on’ when they are 1.
Dummy variables can be used either as explanatory variables or as the
dependent variable.
When they act as the dependent variable there are specific problems with how
the regression is interpreted, however when they act as explanatory variables
they can be interpreted in the same way as other variables.
 CONT…
These are: nominal, ordinal, interval and ratio scale variables.
regression models do not deal only with ratio scale variables; they can also
involve nominal and ordinal scale variables.


Where Yi = the (average) salary of public school teachers in state I, D 1i = 1 if

the state is in the Northeast, 0 otherwise (i.e. in other regions of the country)
D2i = 1 if the state is in the South, 0 otherwise (i.e. in other regions of the
country)
Dummy variable regression is that if there are m categories, we need only m-1
dummy variables.
CONT…
Dummy variables that represent a change in policy:
 Intercept dummy variables, that pick up a change in the intercept of the regression
 Slope dummy variables, that pick up a change in the slope of the regression

Suppose we want to analyze the effect of sex (D)on farm productivity of

maize (y) and
Di = 1 if the household head is male

Di = 0 if the household head is female

yi    Di  ui
 CONT…
We can model this in the following way:
yi    Di  ui
This produces an average maize farm productivity for female hhh of E(y/D i =0)
=.
 The average maize farm productivity of male hhh will be E(y/D i = 1) =  + .
 If sex has a significant effect on maize farm productivity, this suggests that male
hhh have higher maize farm productivity than females.
CONT…

When we have a single dummy variable, we have information for both

categories in the model
Also note that Male = 1 – Female
Thus having both a dummy for male and one for female is redundant.
As a result of this, we always omit one category, whose intercept is the
model’s intercept ( or female).
This omitted category is called the reference category
CONT…

So far we have been dealing with variables that we can measure in quantitative
terms.

However, there are cases where certain variables of great importance are
qualitative in nature.

For instance, we may believe that the level of aggregate consumption

expenditure depends not only on disposable income, but also whether or not the
country is in a period of war or peace.

During war- time we expect consumption to be low as compared to peace-

time.
 CON’T…
One approach to this problem would simply be to estimate two separate
consumption functions and obtain two consumption equations.

There is however, a more efficient procedure involving the estimation of only one
equation if we are willing to make certain assumptions.

Suppose that we hypothesize that war time controls do not alter the marginal
propensity to consume out of disposable income, but instead simply reduce the
average propensity to consume.
CON’T…
By this we mean that the slope remains the same, whereas the constant term
becomes smaller for war- time case.

With this assumption, the consumption function becomes

C t  b0  b1Ydt  b2 Dt  u t , t  1, 2, ..., n,
CON’T…
Where Dt = 0 during peace time years
= 1 for war years

Equation above says that during peace time, when Dt = 0, we have

Ct = b0 + b1Ydt+ut
Which in period of war (Dt=1) becomes

Ct = (b0 + b2) + b1Ydt+ ut

 CON’T…
Suppose the time period under consideration has both war and peace periods.

Using the data, we could estimate the values of the coefficient in equation with
our standard multiple regression equation.
 CON’T…
Suppose that we in fact did this and obtained the equation

ˆ  40  0.9Y  30 D
C t dt t

Let us say that the t- ratio corresponding to the Dt was of sufficient size to
suggest that the parameter b2 is not zero.

We would then conclude that the war had a significant negative effect on
consumption expenditures. The estimated consumption function would be
CON’T…

 Cˆ t  40  0.9Ydt , for years of peace

 Cˆ t  10  0.9Ydt , for war years

If consumption expenditures are measured in billions of dollars, a

comparison of the above two equation would then suggest that, for
corresponding levels of income, consumption expenditures were 30
billion dollars less during years of war.
CON’T…
The use of dummy variables is an extremely powerful extension of regression
analysis.

It allows us to expand the scope of our analysis to encompass variables that
we cannot measure in quantitative terms.

With dummy variables, we can take account of the effects of important

qualitative factors that influence the values of our dependent variable.
CON’T…

We can use as many dummy variables as we like as independent

variables in a regression equation, providing that we have a
sufficient number of observations to allow us to estimate the
equation.
But care should be taken to avoid dummy variable trap.
For the qualitative variable having s categories, we choose one of
them as the omitted category (without loss of generality, category 1)
and define dummy variables D2, ..., Ds for the rest.

What is dummy variable trap?

The coefficient of each dummy variable represents the increase in

the intercept relative to that for the basic category. But there is no
basic category for such a comparison creates dummy variable trap.
CON’T…

Observation Category X1 D1 D2 D3 D4
1 4 1 0 0 0 1
2 3 1 0 0 1 0
3 1 1 1 0 0 0
4 2 1 0 1 0 0
5 2 1 0 1 0 0
6 3 1 0 0 1 0
7 1 1 1 0 0 0
8 4 1 0 0 0 1

June 12, 2024 PREPARED BY K. S. 175

CON’T…
Mathematically, we have a special case of exact multicollinearity. If there is no
omitted category, there is an exact linear relationship between X1 and the dummy
variables.
X1 is the variable whose coefficient is b1. It is equal to 1 in all observations.
Usually we do not write it explicitly because there is no need to do so.

If there is an exact linear relationship among a set of the variables, it is

impossible in principle to estimate the separate coefficients of those variables.
To understand this properly, one needs to use linear algebra.
 CON’T…
If you tried to run the regression anyway, the regression application should
detect the problem and do one of two things. It may simply refuse to run the
regression.

Alternatively, it may run it, dropping one of the variables in the linear
relationship, effectively defining the omitted category by itself.

There is another way of avoiding the dummy variable trap. That is to drop
the intercept (and X1). There is no longer a problem because there is no
longer an exact linear relationship linking the variables.
UNIT 5: ECONOMETRIC
PROBLEMS

Assumptions Revisited
Non-normality
Heteroscedasticity
Autocorrelation
Multicollinearity
 ASSUMPTIONS REVISITED

two major problems arise in applying the classical linear regression model.

1. those due to assumptions about the specification of the model and about
the disturbances and
2. those due to assumptions about the data
 ASSUMPTIONS

 The regression model is linear in parameters.

 The values of the explanatory variables are fixed in repeated sampling (non-
stochastic).
 The mean of the disturbance (ui) is zero for any given value of X

i.e. E(ui) = 0
 The variance of ui is constant i.e. homoscedastic
 There is no autocorrelation in the disturbance terms
 The explanatory variables are independently distributed with the ui.
 CON’T…
 The number of observations must be greater than the number of explanatory
variables.
 There is no linear relationship (multicollinearity) among the explanatory
variables.

 The stochastic (disturbance) term ui are normally distributed

i.e., ui ~ N(0, ²)

 The regression model is correctly specified

i.e., no specification error.

 VIOLATIONS OF ASSUMPTIONS

The Zero Mean Assumption i.e. E(ui)=0

 If this assumption is violated, we obtain a biased estimate of the intercept

term.

 But, since the intercept term is not very important we can leave it.

 The intercept term does not also have physical interpretation.

 The slope coefficients remain unaffected even if the assumption is violated.

 Non-normality

 The OLS estimators are BLUE regardless of whether the ui are normally
distributed or not.

 In addition, because of the central limit theorem, we can argue that the test
procedures:

 The t-tests and F-tests - are still valid asymptotically, i.e. in large sample.
 HETEROSCEDASTICITY: THE ERROR
VARIANCE IS NOT CONSTANT

If the error terms in the regression equation have a common variance i.e., are
Homoscedastic. If they do not have common variance we say they are
Heteroscedastic.

The basic questions to be addressed are:

What is the nature of the problem?

What are the consequences of the problem?
How do we detect (diagnose) the problem?
What remedies are available for the problem?
 THE NATURE OF THE PROBLEM

 In the case of homoscedastic disturbance terms, spread around the mean is

constant.

 But in the case of heteroscedasticity disturbance terms, the variance changes

with the explanatory variable.

 The problem of heteroscedasticity is likely to be more common in cross-

sectional than in time-series data.
 CAUSES OF HETEROSCEDASTICITY

 Following the error-learning models, as people learn, their errors of

behavior become smaller over time thereby the var/se as well.

 As income grows people have discretionary income and hence more scope
for choice about the disposition of their income. Hence, the variance of the
regression is more likely to increase with income.

 Improvement in data collection techniques will reduce errors (variance).

 CON’T…

 Existence of outliers might also cause heteroscedasticity.

 Misspecification of a model can also be a cause for heteroscedasticity.
 Skewness in the distribution of one or more explanatory variables included in
the model is another source of heteroscedasticity.

 Incorrect data transformation and incorrect functional form are also other
sources.
CONSEQUENCES OF
HETEROSCEDASTICITY
 If the error terms of an equation are heteroscedastic:
 estimators are still linear.
 The least square estimators are still unbiased.
But……… there are three major consequences.
 It does affect the minimum variance property.
 The OLS estimators are inefficient.
 Thus the test statistics – t-test and F-test – cannot be relied on in the face of
uncorrected heteroscedasticity.
 DETECTION OF
HETEROSCEDASTICITY
 There are no hard and fast rules (universally agreed upon methods) for
detecting the presence of heteroscedasticity.

 But some rules of thumb can be suggested.

 Most of these methods are based on the examination of the OLS residuals,
 There are informal and formal methods of detecting heteroscedasticity.
 CON’T…

1. Nature of the problem

 heteroscedasticity is the rule rather than the exception.
2. Graphical method
 The squared residuals can be plotted either against Y or against one of the
explanatory variables.
 If there appears any systematic pattern, heteroscedasticity might exist.

These two methods are informal methods.

 CON’T…

3. Park Test- Park suggested a statistical test for heteroscedasticity based on

the assumption that the variance of the disturbance term (i²) is some
function of the explanatory variable Xi.

 d i  test
 Correlation
2
4. Spearman’s Rank
rS  1  6 
 N ( N  1) 
2

Where d is difference between ranks

CON’T…

5. Goldfeld and Quandt Test

 This is the most popular test and usually suitable for large samples.
 It is assumed that if the variance (i²) is positively related to one of the
explanatory variables in the regression model and if the number of observations
is at least twice as many as the parameters to be estimated, the test can be used.

Yi   0   1 X i  U i

 i2   2 Xi2
REMEDIAL MEASURES

Remedies for Heteroscedasticity

 Include a previously omitted variable(s) if

heteroscedasticity is suspected due to omission of
variables.

 Redefine the variables in such a way that avoids

heteroscedasticity. For example, instead of total
income, we can use Income per capita.
AUTOCORRELATION:ERROR TERMS ARE
CORRELATED

 Is a problem which occurs when the assumption “non-existence of serial

correlation (autocorrelation) between the disturbance terms, Ui’ is violated.

Cov(U i , V j )  0 i  j

 Serial correlation implies that the error term from one time period depends
in some systematic way on error terms from other time periods.
 Autocorrelation is more a problem of time series data than cross-sectional
data.
 If by chance, such a correlation is observed in cross-sectional units, it is
called spatial autocorrelation.
 CAUSES OF AUTOCORRELATION

 Inertia or sluggishness in economic time-series is a great reason for autocorrelation.

 For example, GNP, production, price index, employment, and unemployment exhibit
business cycles.

 In this upswing, the value of a series at one point in time is greater than its previous
values. These successive periods (observations) are likely to be interdependent.
 Specification bias – exclusion of important variables or incorrect functional forms
 Lags – in a time series regression
 Manipulation of data – if the raw data is manipulated (extrapolated or interpolated)
 CONSEQUENCES OF SERIAL CORRELATION

 The estimates of the parameters remain unbiased even in the presence of

autocorrelation but the X’s and the u’s must be uncorrelated.

 Serial correlation increases the variance of the OLS estimators.

 The variance of the disturbance term, Ui may be underestimated.

 If the Ui’s are autocorrelated, then prediction based on the OLS estimates will
be inefficient.
 DETECTING AUTOCORRELATION

1. The existence of autocorrelation may be gained by plotting the residuals

either against their own lagged values or against time.
2. Durbin-Watson Test
This test is appropriate only for the first order autoregressive scheme.
 CON’T…
This test is applicable if the underlying assumptions are met:
The regression model includes an intercept term
The serial correlation is first order in nature
There are no missing observations in the data

The equation for the Durban-Watson d statistic is

 (e t  et 1 ) 2
d  t 2
N

e
2
t
t 1
 REMEDIAL MEASURES FOR AUTOCORRELATION

1. The solution depends on the source of the problem.

If the source is omitted variables, the appropriate solution is to include
these variables.
If the source is misspecification of the mathematical form the relevant
approach will be to change the form.
2. If these sources are ruled out then the appropriate procedure will be to
transform the original data.
 MULTICOLLINEARITY

 It exists when there is exact linear correlation between Regressors

 It appear when the assumption ‘no independent variable is a perfect linear
function of one or more other independent variables’ is violated
 Multicollinearity is not a condition that either exists or does not exist in
economic functions, but rather an inherent phenomenon in most relationships
due to the nature of economic magnitude.
 Serious if VIF is greater than 10
 CONSEQUENCES OF
MULTICOLLINEARITY

 The estimates of the coefficients are statistically unbiased.

 When multicollinearity is present in a function, the variances and therefore
the standard errors of the estimates will increase, although some
econometricians argue that this is not always the case.
 The computed t-ratios will fall i.e. insignificant t-ratios will be observed in
the presence of multicollinearity.
 A high R² but few significant t-ratios are expected in the presence of
multicollinearity.
DETECTING MULTICOLLINEARITY

 High R² but few significant t-ratios

 VIF and Tolerance test

 High pair-wise (simple) correlation coefficients among the repressors

(explanatory variables).
REMEDIES FOR
MULTICOLLINEARITY
 Do Nothing
 Dropping one or more of the multicollinear variables
 Transformation of the variables
Two common transformations are:

to form a linear combination of the variables

to transform the equation into logs
 Increase the sample size
 Other Remedies (Factor analysis and Principal component analysis or
other techniques such as ridge regression).
Any Q uestion …

Q
GOOD LUCK !!!