Correlation Vs. Causation

Correlation is very often misunderstood as causation i.e. a cause and effect relationship. But as a matter of fact, correlation implies only convariation between any two, or more variables. A very high degree of correlation obtained from the calculation does not necessarily mean that there is some cause and effect relationship between the two variables. A correlation measure may give us some value of co-efficient of correlation between the variables of marks and weight but it cannot be concluded there from that the ‘mark’ variable can be a cause, or effect of the weight variable under any circumstances. Such type of conclusion, or effect of the weight variable under any circumstances. Such type of conclusion, or interpretation of cause and effect relationship is nothing but non-sense, or spurious. Therefore, before interpreting the value of correlation between any two variables as the causation, or the cause and effect relationship, care must be taken to see that the two variables are of such nature that there can exist some sort of relationship between them in reality for which one of them can be either a cause, or an effect of another.

In this connection, the following points should be taken into consideration before interpreting the correlation as causation.

• The correlation between any two series may be observed due to pure chance as under:

MarksHeights

2012

3018

4024

5030

6036

7042

80 48

• Both the variables may be affected by a third variable. For example, increase in the amount of rainfall may increase the yield of both rice and tea which may appear that the increase in the yield of rice is either a cause, or effect of the increase i8n the yield of tea.
• Both the variables may be mutually affecting each other. For example, the increase in price causing decrease in demand, and increase in demand causing increase in price. Here, both the price variable, and the demand variable affect each other.