Two Correlation Coefficients You May Not Have Heard

Author(s): Albert Wibowo

Originally published on Towards AI.

Photo by Alina Grubnyak on Unsplash

The concept of correlation measure is one of the most fundamental concepts in statistics. It can be understood easily and is very useful. But, it may take a long time to master. It goes deeper than the phrase correlation does not imply causation. In doing our analysis for example, we must be mindful of things such as:

Reverse causality — instead of A causes B, B causes ACommon causal variable — C causes both A and B, giving the illusion there is a cause and effect between A and BBidirectional causation — A causes B and B also causes A

The properties of the variables themselves can also add an extra layer of complexity to the analysis. The properties include things such as:

The type of the variable — continuous, categorical, and ordinalThe interaction with other variables — monotonic and non-monotonic

The current de facto standard to measure the linear correlation between two continuous variables, for example, will be Pearson’s correlation, while Cramer’s V or Chi-squared test can be used to calculate the correlation between two categorical variables. In the case of ordinal variables, we can use either Spearman’s or Kendall’s Tau correlation.

While the so-called de facto methods are… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI