Page 245 - DSOC404_METHODOLOGY_OF_SOCIAL_RESEARCH_ENGLISH

Unit-29: Correlational Analysis: Test of Significance and Co-variance

29.3 Kinds of Correlation notes

Kinds of correlation can be described in two different ways-
1. positive and negative correlation. According to Prof. Kataria, “When the two variables change in
one direction then the correlation is said to be direct or positive. for example, if the price of a substance
increases and the supply of the substance also increases then exists a positive correlation between
them.” the same thing has been described by Prof. elhance in different words. He writes, “When the
price of two variables increases or decreases in the same direction, for example, the increase in the
price of any variable is related to the increase in the price of other variable and the decrease in the
price of variable is related to the decrease in the price of other variable than the correlation is said to be
positive. in opposite of this if the price of two variables increase or decrease in opposite direction, like
the increase in price of variable is related to the decrease in the price of other variable and in the same
way the decrease in the price of a variable is related to the increase in the price of another variable then
the correlation is said to be negative.” in other words, “if two variables change in opposite direction
instead of same direction then their correlation is said to be indirect or negative.”

Did You Know? the increase in the price of a substance results in a decrease in its demand.
in this way demand and price have a negative correlation.

2. linear and non-linear correlation. explaining the meaning of these Prof. elhance writes, “When
there is a constant ration between the prices of two variables then it is said to be linear correlation. i.e.,
if every time the price increases by 10% then the increase in supply by 20% proves linear relationship
between the two. This can be shown in the form of a straight line. In financial and social data such
type of relationship very rarely occurs, specifically in social facts the ratio between the changes in
two variables is rarely constant. thus there change cannot be shown on a straight line. this type of
correlation is said to be curvilinear or non-linear correlation.

Notes in the words of Prof. elahance, “Linear correlations are those in which there is a
constant ratio between the related variables and non-linear correlations are those
where such ratio increase and decrease.”

29.4 Methods of Measurement of Correlation

the correlation between two or more categories can be found out with following techniques-
(i) scatter Diagram
(ii) Correlation Graph
(iii) Coefficient of Correlation
(iv) Correlation table

29.5 scatter Diagram or Dotogram

In this correlation is presented in the form of figures but the correlation is not in the form of exact
numbers rather it is in the form of estimate. the technique to make this is similar to linear point

loVely professional uniVersity 239