Scribd
Upload
268 views

Regression

The document discusses regression analysis using sales data from a company. It provides the regression equation relating profit to sales as Y = 1.081 + 1.4527X. There is a very high positive correlation (r = 0.98) between sales and profit, with sales explaining 96.1% of the variation in profit. When sales are 50 lakh, the estimated profit is 73.731 thousand. The document also analyzes a multiple regression on sales revenue using cost of production and number of advertisements as independent variables.

Uploaded by

oove
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
268 views

Regression

The document discusses regression analysis using sales data from a company. It provides the regression equation relating profit to sales as Y = 1.081 + 1.4527X. There is a very high positive correlation (r = 0.98) between sales and profit, with sales explaining 96.1% of the variation in profit. When sales are 50 lakh, the estimated profit is 73.731 thousand. The document also analyzes a multiple regression on sales revenue using cost of production and number of advertisements as independent variables.

Uploaded by

oove
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 6

CORRELATION AND REGRESSION

ANALYSIS
Regression Analysis: An Example

• Data on sales and profit of a company


are given below:
X Y
Sales (Lakh Profit (Thousand
Tk.) Tk.)

10 15
12 20
14 20
16 25
20 30

i. Find the equation of regression of profit on sales


ii. Estimate the relationship between sales and profit in
relative term
iii. Assess the explanatory power of the independent
variable
iv. Estimate the profit when the volume of sales is Tk. 50
lakh
Interpret the results

Solutions:

i. The required equation is :


Y = a + b X, where

Y – Profit (Dependent/Criterion Variable)


X – Sales (Independent/explanatory/predictor
variable)
a - Constant/Intercept
1
b – Regression coefficient

we use the following formula:

ΣXY - ΣX.ΣY/n
b= --------------------
ΣX2 - (ΣX)2/n

ΣY - bΣX
a= -------------
n
___ ___
Or
a=Y–bX

X Y XY X2 Y2
10 15 150 100 225
12 20 240 144 400
14 20 280 196 400
16 25 400 256 625
20 30 600 400 900
72 110 1670 1096 2550

We get, b= 1.4527

Putting the value of b in equation (1) we get a = 1.081

Hence the desired equation is: Y = 1.081 + 1.4527 X

Interpretation:
If the volume of sales increases for Tk. 1 lakh, profit of the
company increases for Tk. 1.4527 Thousand.
ii. The relationship between the variables in relative
term can be estimated with the help of the coefficient
of correlation, r.

2
ΣXY - ΣX.ΣY/n
r= ----------------------------------
_______________________
V [ΣX2 - (ΣX)2/n] [ΣY2 - (ΣY)2/n]

r= 0.98 –

Interpretation:

r= 0.98 – indicates a very high degree of positive relationship


between sales and profit

iii. The explanatory power of the independent variable


can be assessed by the coefficient of determination
(r2), r2 = 0.961

Interpretation:

r2 = 0.961 – indicates that 96.1% of the variation in the profit


can be explained by the variation in the sales
Or, Volume of sales explains 96.1% of the variation in the profit.

iv. When X= Tk. 50 lakh, Y=73.731 thousands

Y= 1.081 +1.453*50=73.731 thous Tk.

Interpretation:
When the volume of sales is Tk. 50 lakh, profit will be Tk.
73.731 thousands.

3
6 Brittle Biscuits are new in Bangladesh. The Sales Manager of the
company has seen that in the last few months the sales revenue has
fluctuated widely. She was able to point fingers to two factors
affecting the sales revenue. Number of advertisements played on TV
and radio was one of them. The other factor was the type of flour
used in producing the biscuits, which, in turn, affected the cost of
production. Results of the multiple regression equation of sales
revenue on cost of production and no. of advertisements are given
below:
(a) Find the equation of regression of sales revenue on cost of
production and no. of advertisements.
(b) Estimate the relationship among the variables in relative
terms
(c) Assess the explanatory power of the independent variables
(d) Indicate which of these variables exerts more influence on
sales revenue.
(e) Assess the significance of the results
(f) Ascertain whether there is a problem of multicollinearity.
Interpret the results.
Regression
Variables Entered/Removed(b)
Model Variables Entered Variables Removed Method
ADV: No. of Advertisements,
1 . Enter
CP: Cost of Production (in Lac Tk.)
Dependent Variable: SR- Sales Revenue (in Lac Tk.)

Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .995 .991 .988 1.1386

ANOVA
Model Sum of Squares Df Mean Square F Sig.
1 Regression 947.325 2 473.662 365.345 .000

4
Residual 9.075 7 1.296
Total 956.400 9

5
Coefficients(a)
Unstandardized Standardized
Coefficients Coefficients
t Sig.
Beta
Model B Std. Error

-
(Constant) -15.658 5.639 .027
2.776
1
CP 3.235 .905 .769 3.575 .009
ADV .713 .669 .229 1.066 .322
a Dependent Variable: SR

Correlation Matrix

SR CP ADV

SR 1.000 .994 .987


Pearson Correlation CP . 1.000 .985
ADV 1.000

You might also like