Correlation or causality

Francis Galton, Darwin’s cousin, best known for promoting eugenics, invented the mathematical concept of correlation. Since then, many mistakes have been made relating correlation and causality, which sometimes coincide, but can also be completely different.

Two variables are said to be correlated when increases in one resemble increases in the other, and decreases in one resemble decreases in the other. But not all correlations are the same: to distinguish them, Galton devised the correlation coefficient, a number between -1 and 1.

• Two series of numbers (whatever their origin) have a correlation coefficient equal to 1, if the increases and decreases of the two series are exactly the same.
• The coefficient will be -1, if the increases in one are exactly the same as the decreases in the other and vice versa.
• The coefficient will be 0 if there is no relationship between the increases and decreases of the two series.
• The correlation will be strong if its coefficient is greater than 0.5.
• The correlation will be weak if we have a value between 0 and 0.5.
• The inverse (or negative) correlation will be strong if we get a negative value between -0.5 and -1.
• The inverse correlation will be weak if we have a negative value between 0 and -0.5.

Human beings have a great tendency to confuse correlation with causation. When two situations are correlated, we almost automatically think that one of them is the cause of the other. The application of this rule can give rise to a fallacy known by the Latin name cum hoc ergo propter hoc, which means both things go together, therefore one is the cause of the other.

When two variables A and B are strongly correlated, the following situations may occur:

1. A is the cause of B.
2. B is the cause of A.
3. A and B are related through a phenomenon C, which could be a common cause of A and B, or have some other relationship with them.
4. The strong correlation is pure chance.

To see that case 4 is real,  we can look at those compiled in this article. Or these two, which I have found after a Google search: 

• U.S. spending on science, space and technology is strongly correlated with suicides by hanging, strangulation and suffocation. 
• US oil import from Norway is strongly correlated with the number of drivers killed in a train collision. 

It is evident that all these cases are pure coincidence. It would be easy to prove it. Just wait a while and add some more data to both series, which will surely stop being correlated, because a casual correlation is unlikely to last for long.

On June 5, 2024, Madri+d echoed a scientific news item presented under the following headline:

A large meta-analysis links global warming to a significant increase in suicides

The text of the article assumes that global warming is the cause of the increase in the number of suicides, i.e. option number 1 above has been chosen. What can we think?

Obviously, case 2 can be eliminated: the number of suicides is NOT the cause of global warming. Due to the large amount of data used in the meta-study, case 4 can probably be eliminated as well. But, if case 3 is true, there could be an indirect causal relationship between global warming and the increase in the number of suicides. Perhaps this could be due to the imposition of pernicious ideologies throughout the world, which leads many people to lose the wish to go on living; especially now, when many young people are hooked on social networks and do not develop direct human relationships. Global warming could play a role, through the feeling of anguish caused by the sometimes excessively alarmist forecasts published by the media and disseminated by many educators.

Let me state here what I think about this question:

1. Scientific data indicate that we are suffering from an anthropogenic climate change with a speed unparalleled in the history of the Earth.
2. As I said in another post, prior to the preceding observation, it is necessary to do something to avoid the worst consequences of climate change.
3. But an important cause of the increase in the number of suicides could be the fact that certain political and economic bodies are taking advantage of this situation to manipulate and control the population, to eliminate democracy and freedom, and to impose their ideology.

The same post in Spanish

Thematic Thread on Politics and Economics: Previous Next

Thematic Thread on Mathematics and Statistics: Previous Next

Manuel Alfonseca