By Dennis John, May 2020
Nowadays, correlation and causation are two terms which are often conflated into one. Correlation is the term used to define a statistical measure which measures how linked two variables are with each other. It is measured on a scale from 0 to 1, with 0 suggesting there is no relationship between the two variables, and 1 suggesting there is a perfect relationship between them. Meanwhile, causation indicates that one event directly leads to another event occurring.
It is important to realise that two variables correlating does not mean that one causes the other. For example, over the last ten years, both car ownership levels and average life expectancies have risen across the world. However, that does not mean that one of those variables caused the other. Although there is correlation between the two variables, it is very unlikely that one of them caused the other. In fact, they are both probably influenced by a third variable, which is often known as a confounder. In this case, over time, there has been increased spending on innovations in healthcare and technology, resulting in better healthcare facilities and better factories, which result in increased life expectancies across the world, and increased car production, respectively.
However, it is not always so easy to distinguish between correlation and causation. In the 1920s, researchers went to the Hawthorne Works in Cicero, Illinois to investigate the effect on labour productivity levels by adjusting working conditions. Although there was a marked improvement in labour productivity levels in the factory, as soon as the researchers left, productivity levels dropped back to where they were before. In this case, the two variables are labour productivity levels and working conditions. Although there seemed to be a positive correlation between the variables, the researchers believed, in truth, that the increased productivity was as a result of researchers being interested enough in their working conditions to conduct an experiment on it. This phenomenon is called the Hawthorne Effect.
However, in my opinion, the main problem with correlation and causation is a lot worse. We currently live in a data-driven world. Unfortunately, statistics can often be misconstrued to portray false information to the general public, and the difference between causation and correlation can often be manipulated to significantly impact outcomes in situations. On a more global scale, politicians tend to use this tactic quite frequently. Although he never used any graphs explicitly, Donald Trump continuously attempted to support his ban on Muslim-majority countries by saying that there was an increase in crime in the US, as well as an increase in the US’s Muslim population, without ever having any evidence to make one believe that the Muslims were the ones committing the crimes. This did, unfortunately, lead to an increased anti-Muslim sentiment from the residents of the US. Despite the fact that there was a correlation between the increase in crimes and the Muslim population, there was never any conclusive proof that Muslims caused the rise in crimes.
Comments