How do you know if a correlation is spurious?
How do you know if a correlation is spurious?
In 1897, Karl Pearson suggested the theory of spurious correlation, which describes how a valid value for a coefficient of correlation can be obtained when the two variables are completely unrelated. A spurious correlation occurs when two variables are statistically related but not causally related. These two variables are often incorrectly linked to each other due to an unknown third factor.
When there is indication of relationship between variables, but the correlation does not indicate a causal effect from one variable to the other, it is referred to as "spurious correlation." Only when variables are actually correlated, i.e., statistically linked and therefore statistically not independent, does the term "spurious correlation" have any significance.
Additional variables and equations must be added, and adequate hypotheses must be made to define the conditions of this larger structure, in order to evaluate whether a correlation between two variables is true or misleading. The association is "valid" if the two original variables are causally linked in the given context.
Step by step
Solved in 3 steps
