Correlation Initially developed by Sir Francis Galton (1888) and Karl Pearson (1896)

Sir Francis Galton 1822-1911

What is the Need to study correlation

Correlation help in study economic theory and business studies, it help in establishing relationship between variable like price and quantity demanded, advertising and sales promotion measures Correlation analysis helps in deriving precisely the degree and direction of such relation. The effect of correlation is to reduce the range of uncertanity of our prediction. The prediction based on correlation analysis will be more reliable and near to reality. The measure of coefficient of correlation is relative measure of change.

On the basis of direction

Positive correlation: if both the variables vary in same direction, correlation is said to be positive correlation .if one variable increase, the other also increases or, if one variable decreases, the other variable is said to be a positive correlation.

Positive correlation

Negative correlation: if both the variable vary in opposite direction, the correlation is said to be negative. In other word if one variable increases, but other variable decreases or, if one variable decreases but the other variables increases, than correlation between two variables is said to be negative correlation.

Negative correlation

On the basis of ratio of changes in variable Linear correlation: If the amount of changes in one variable bears a constant ratio to the amount of changes in the other variable, then correlation is said to be linear. if such variable plotted on graph paper all the plotted points would fall on straight line.

Non-linear correlation : If change in one variable does not bear a constant ratio to the amount of changes in the other variable, then correlation is said to be nonlinear. if such variable plotted on graph, the point will fall on a curve and not a straight line.

Methods of studying correlation

Graphic 1.Scatter diagram Algebric
1.Karl parson's coefficent of correlation

2.Graphic

method

2.Rank metho d

Scatter diagram method Scatter diagram is diagrammatic representation of bivariant data to ascertain the correlation between two variable. Perfect positive correlation

Perfect negative correlation

High degree positive

High degree negative

No correlation

Graphic method

Value of r (a) If r = +1 (b) If r =-1

interpretation Perfect positive correlation Perfect negative correlation

(c) If r =0

No correlation

(d) If +.75 r < +1

High positive correlation

(e)If -.75 r > -1

High negative correlation

(f)If + 0.50 r < 0.75

Moderate positive correlation

(g)If -0.50 r > -0.75

Moderate negative correlation

(h)If r < +0.50 (i)If r> -0.50

Low positive correlation Low negative correlation

Karl Pearson method

We have seen that the scatter diagram method helps us in knowing the direction of the relationship between the variables but it can not give us exact amount of the relationship between the variables. So for knowing exact amount of relationship between two variable Karl Pearson method is more useful.

Properties of correlation co-efficient:

1) The correlation co-efficient lies between -1 and +1. 2) The correlation co-efficient is independent of change of origin and scale. 3) The correlation co-efficient is an absolute number and it is independent of units of measurement.

4) r2 always lies between 0 and 1. Merits of Karl Pearson's co-efficient of correlation

1.)correlation coefficient gives direction as well as degree of relationship between the variables.

2.)correlation coefficient along with other information help in estimating the values of the dependent variable from the known value of independent variable.

Limitation of Karl Pearsons co-efficient of correlation

1.)Assumption of linear relationship:-The assumption of 2.)Time consuming :- Its computation is time compared to other method . 3.)Affected by extreme values: -It is affected by the value of extreme item. 4.)Requires careful interpretation:- The investigation should reach a conclusion based on logical reasoning and investigation on significantly related matter. linear relationship between the variables may or may not hold always true . consuming as