Question

1.      Clearly explain the difference between correlation and causation. Explain which type of connection is harder to prove and why. Provide two clear examples to illustrate the differences between the two concepts.

Answer 1

The term "correlation" is made of two words "co", means between or among, and "relation" means the relationship or association. The correlation, therefore is the relationship between two or more variables considered to be related in statistical context. It indicates the change in value of one variable (increases or decreases), is related to the value of other variable ( it although may be in different direction). For example: Increase in temperature, increase in sales of ice-cream ( positively correlated- in same direction). Similarly, increase in temperature, decrease in sales of woolen clothes ( negatively correlated-in different direction).

While the term "causation" indicates the cause or reason behind occurrence of an event due to presence of another event. It seeks to provide a causal relationship between two variables, one as cause and another as effect. for example: smoking causes cancer.

Out of both the connections, causation is more difficult to prove, as although statistically causation can be established but in practice it is difficult to clearly establish cause and effect, as compared to establishing correlation. For example, in a study it is found that sale of ice-cream was positively correlated with the police reporting of road rage incidences, but that does not mean that ice-cream was cause behind road rage. Instead it was found that it was due to presence of another mediating variable of summer season that lead to road rage.

Therefore, correlation may prove nothing but the relation between two variables while causation between variables may or may not be there, due to presence of external variables or mediating variables.