BUSINESS STATISTICS NOTES (UNIT 3)

cORRELATION
CORRELATION- This is defined as relationship between two or more
variables are independent. variables. In correlation both
According to A.M. Tuttle " correlation is an analysis of the
covariation between two or more variable".
TYPES OF CORRELATION- Correlation is
classified on three basis
ON THE BASIS OF
DIRECTION
1 Positive correlation
2 Negative correlation
ON THE BASIS OF CHANGE IN
PROPORTION
3 linear correlation
4 Non linear correlation
ON THE BASIS OF NO OF
VARIABLES
5 Simple correlation 6 Partial correlation
7Multiple correlation
Positive colarretion - If values of
variables are varying (increasing or
positive or direct correlation.decreasing)
direction then such correlation is called in the same

Examles 1 X:5 10 15 17

Y:10 12 16 18 20

Example2- income and expenditure

Negative colarretion- If values of variables are
direction then such correlation is called varying (increasing or decreasing) in the
negative or inverse correlation. opposite
Examle1 X: 17 15 10 5

Y: 2 7 13 14

Example 2 price and demand

Linear correlation -A correlation is said to
linear if a constant change in one of the
with respect to a change in the variable values
corresponding values of another variable.
Example 1 X: 10 20 30 40 50

Y: 40 60 80 100 120

Example 2 cost and profit

 2
Non linear correlation -A
correlation is said to be a non linear when the amount of
values of one variable does not bear a change in the
constant ratio to the amount of change in
values of another variables. corresponding
Example1 X: 9 10 10 28 29 30
Y: 80 130 70 150 230 560 460 600
Examle 2 sale and profit
Simple correlation -A correlation is said to be
simple if only two variables are involved and the
relationship is studied between thease two variables.
Partialcorrelation-A Correlation is said to be partial if
the relationship is studied more than two variables are involved but
between two variables keeping other variables as
constant.
Multiple correlation - ACorrelation is said to be
but relationship is studied more than
multiple if more than two variables are involved
two variable simultaneously.
Example Production of wheat depends on so many
factors like rainfall, quality of
if we study the relation
between production of wheat per hectare and quality of seed, manure etc.
rainfall and manure as constant then correlation is seed, keeping
partial if we study the relationship between
production of wheat per hectare and quality of seed, rainfall,
correlation is multiple. manure simultaneously then

REGRESSION
Regression- The statistical method that helps to formulate an
or more variables in the form of an alzebraic relationship between two
equation to estimate the value of random variable,
value of another variable is called given.the
regression.
According to Blair " Regression is the measure of
variables in terms of the original units of data." average relationship between twO or more
Note - In regression analysis there is only one
both variable have to be dependent variable (unknown) while in correlation
independent (known).
Types of regression -there are following
important types of regression
1Simple and Multiple Regression -if a regression model
dependent variable and only one independent variable then characterizes the relationship between a
such a regression model is called
but if more than one independent
variable are associated with dependent variable then such asimple
-regression model is called multiple regression.
Example Sales turn over of a product (dep variable) is
such as price of the pYoduct, expand on associated with multiple independent variable
advertising
consider sales turn over with respect to priceof the ,quality of the product, competitors so on. If we
product then such regression is simplebut if we
consider sales turn over with respect to price of the product, quality of the
product,expand on
advertising then such regression is multiple.
2 Linear and Non linear Regression-A Linear regression analysis is one which gives rise to a
Straight line when the data relating to the two variables are ploted on graph paper. On
other hand a
non linear regression analysis is one which gives rise to a Curved line
when data relating to two
variables are ploted on graph paper.
EQUATIO OF LINEAR REGRESSION Y=a + bx

EQUATION OF NON LINEAR REGRESSION Y=a +bx +c x*

Shows Limear qnd ton drea

The Falowng Graph
Regreson

Linear Regrei on
Hon iner Pegrnin
2.
 Rege9on

Regresion Meahy Stebpina bch toards the

Satutics Yeyreion amelis
is atticatle to al thesr
thog fiddy whoe

two mon laled Vaiales hare th.

g bach to the mran,

Accordig o Blair. Rgreuion

4he meaqur veat vdatonghit
bet en two more

data Corelatton an Begyion

Ditterence betwe (n
( ) Condation teu abuut :)Thragh
rlatioghi beten byediction ee timatie
two more Variade.

Vamale Varie is Rhence

Chnon)
hn

Catficlnt (
 n

Vaine
Cotry
Carete
Gueltty

tre