Department of Commerce and Accounting & Finance, Emerson University Multan

EViews and Regression Analysis

Case 1:

A cable wire company has spent heavily on advertisements. The sales and advertisement

expenses (in thousand rupees) for the 12 randomly selected months are given in below table.

Develop a regression model to predict the impact of advertisement on sales.

Month Adv Sales

Jan 92 930
Feb 94 900
Mar 97 1020
Apr 98 990
May 100 1100
Jun 102 1050
Jul 104 1150
Aug 105 1120
Sep 105 1130
Oct 107 1200
Nov 107 1250
Dec 110 1220

Financial Econometrics ® by Dr. Irfan Javaid © 1

 Department of Commerce and Accounting & Finance, Emerson University Multan

Solution:

Step 1: Set Null and Alternative Hypotheses

𝑆𝑎𝑙𝑒 = 𝑓(𝐴𝑑𝑣𝑒𝑟𝑡𝑖𝑠𝑒𝑚𝑒𝑛𝑡)
𝑆𝑎𝑙𝑒𝑡 = 𝛼 + 𝛽𝐴𝑑𝑣𝑒𝑟𝑡𝑖𝑠𝑒𝑚𝑒𝑛𝑡𝑡 +∈𝑡
H0: β = 0 (There is NO linear relationship between sales and advertisement) Reject
H1: β ≠ 0 (There is A linear relationship between sales and advertisement) Accept

Step 2: Determine the appropriate statistical test

The t test is an appropriate test because the sample size is less than 30. The t statistic can

be computed by using the formula:

Step 3: Set the level of significance

Level of significance α is set as 0.05 or 5%

Financial Econometrics ® by Dr. Irfan Javaid © 2

 Department of Commerce and Accounting & Finance, Emerson University Multan

Step 4: Set the decision rule

For 95% confidence level (α= 0.05), the critical value of z is given as z = ±1.96.

If the computed value of z is between +1.96 and –1.96, the decision is to accept the null

hypothesis and if the computed value of z is outside ±1.96, the decision is to reject the null

hypothesis (accept the alternative hypothesis).

Financial Econometrics ® by Dr. Irfan Javaid © 3

 Department of Commerce and Accounting & Finance, Emerson University Multan

Step 5: Collect the sample data

In this stage of sampling, data are collected and the appropriate sample statistics are

computed. The first four steps should be completed before collecting the data for the study.

Note: It is not advisable to collect the data first and then decide on the stages of hypothesis

testing.

Month Advertisement Sales

Jan 92 930
Feb 94 900
Mar 97 1020
Apr 98 990
May 100 1100
Jun 102 1050
Jul 104 1150
Aug 105 1120
Sep 105 1130
Oct 107 1200
Nov 107 1250
Dec 110 1220

Sample size n = 12

Financial Econometrics ® by Dr. Irfan Javaid © 4

 Department of Commerce and Accounting & Finance, Emerson University Multan

Step 6: Analyse the data

At this stage the value is computed from OLS output table from EViews.

Dependent Variable: SALES

Method: Least Squares
Date: 07/04/21 Time: 12:09
Sample: 2020M01 2020M12
Included observations: 12

Variable Coefficient Std. Error t-Statistic Prob.

C -852.0842 203.7759 -4.181477 0.0019

ADV 19.07044 1.999943 9.535496 0.0000

R-squared 0.900917 Mean dependent var 1088.333

Adjusted R-squared 0.891009 S.D. dependent var 112.3981
S.E. of regression 37.10688 Akaike info criterion 10.21649
Sum squared resid 13,769.21 Schwarz criterion 10.29731
Log likelihood -59.29896 Hannan-Quinn criter. 10.18657
F-statistic 90.92568 Durbin-Watson stat 3.005353
Prob(F-statistic) 0.000002

Financial Econometrics ® by Dr. Irfan Javaid © 5

 Department of Commerce and Accounting & Finance, Emerson University Multan

Step 7: Arrive at a statistical conclusion and business implication

𝑆𝑎𝑙𝑒 = 𝑓(𝐴𝑑𝑣𝑒𝑟𝑡𝑖𝑠𝑒𝑚𝑒𝑛𝑡)
𝑆𝑎𝑙𝑒𝑡 = 𝛼 + 𝛽𝐴𝑑𝑣𝑒𝑟𝑡𝑖𝑠𝑒𝑚𝑒𝑛𝑡𝑡 + 𝜖𝑡
̂𝑡 = 𝛼̂ + 𝛽̂ 𝐴𝑑𝑣𝑒𝑟𝑡𝑖𝑠𝑒𝑚𝑒𝑛𝑡𝑡
𝑆𝑎𝑙𝑒
̂𝑡 = −852.0842∗ + 19.07044∗ 𝐴𝑑𝑣𝑒𝑟𝑡𝑖𝑠𝑒𝑚𝑒𝑛𝑡𝑡
𝑆𝑎𝑙𝑒
(203.7759) (1.999943)

Explanation of table:

1. Estimated Coefficients

The calculated value of t-stat is 9.53. The calculated value of t-stat (= 9.53) > tabular

value of t (= 2). Hence, the null hypothesis is rejected and the alternative hypothesis is accepted.

So, it can be concluded there is a significant relationship between two variables. It means, if

there is one unit change (in Rs.) in advertisement, the sale will increase by 19.07044 unit (or Rs.

19.07044).

2. Standard Error of the Coefficients

The standard error of the coefficient measures how precisely the model estimates the coefficient's

unknown value. The standard error of the coefficient is always positive. The smaller the standard

error, the more precise the estimate.

Financial Econometrics ® by Dr. Irfan Javaid © 6

 Department of Commerce and Accounting & Finance, Emerson University Multan

3. Coefficient of Determination:

R2 is calculated as 0.9009. This indicates that 90.09% of the variation in sales can be

explained by the independent variable, that is, advertisement. This result also explains that

9.91% of the variation in sales is explained by factors other than advertisement.

4. Standard Error of the Regression

The value of standard error of the estimate is 37.10688. A large standard error indicates a

large amount of variation or scatter around the regression line and a small standard error indicates

small amount of variation or scatter around the regression line. A standard error equal to zero

indicates that all the observed data points fall exactly on the regression line.

5. Measure of Variations

While developing a regression model to predict the dependent variable with the help of the

independent variable, we need to focus on a few measures of variations.

Total sum of squares (SST) = Regression sum of squares (SSR) + Error sum of squares (SSE)

a. Total variation (SST) can be partitioned into two parts: variation which can be attributed

to the relationship between advertisement and sale; and unexplained variation.

b. The first part of variation, which can be attributed to the relationship between

advertisement and sale is referred to as explained variation or regression sum of squares

(SSR).

Financial Econometrics ® by Dr. Irfan Javaid © 7

 Department of Commerce and Accounting & Finance, Emerson University Multan

c. The second part of variation, which is unexplained can be attributed to factors other than

the relationship between advertisement and sale; and is referred to as error sum of squares

(SSE).

Total sum of squares (SST) = Regression sum of squares (SSR) + Error sum of squares (SSE)

138,966.6667(SST) = 125,197.4582 (SSR) +13,769.20842 (SSE)

6. Log Likelihood Value

Log Likelihood value is -59.2986. It is a measure of goodness of fit for any model (Higher

the value, better is the model).

7. Testing the Overall Model

The F test is used to determine the significance of overall regression model in regression

analysis. The hypothesis that all of the slope coefficients (excluding the constant, or intercept) in

a regression are zero.

H0: all coefficients (except intercept) = 0 (Over all model is NOT fit) Reject
H1: all coefficients (except intercept) ≠ 0 (Over all mode is fit) Accept

The value F-test is 90.92568 and is statistically significant based on probability, it means

the overall regression model is fit.

Financial Econometrics ® by Dr. Irfan Javaid © 8

 Department of Commerce and Accounting & Finance, Emerson University Multan

8. Mean Dependent Variable

The means of the dependent variable, i.e. sale is 1088.33, it is the simple average of

dependent variable.

9. SD Dependent Variable

The standard deviation of the dependent variable, i.e. sale is 112.39.81, it shows the

dispersion from the mean. The greater the value of the dispersion, the greater will be the magnitude

of the deviation from the mean.

10. Adjusted R2, Akaike info criterion, Schwarz criterion and Hannan-Quinn

criterion

All these values are used for the comparison of the model.

1. DV should be same.

2. the value of Adj R2 should increase.

3. the value of AIC/SBC/HQC should decrease.

11. Durbin Watson Stat

The independence of errors is one of the basic assumptions (i.e. auto-correlation among

error terms). The rule of thumb is DW-stat ≈ 2. The DW-stat is 3.005353 which means that there

is NO auto-correlation.

Financial Econometrics ® by Dr. Irfan Javaid © 9