Correlation and Causation Pt 2

Correlation and Causation

At the beginning of my scholarly activities, we were taught ad nauseam a simple phrase “Correlation does not mean causation”.  This simple little phrase helped me on more than one occasion and beyond just a test in a class. For example in 1975 a study in Taiwan looked at contraceptives, they collected data on a wide range of variables and the variable that most strongly related to contraceptive use was the number of electrical appliances in the home. Great, now we can start handing out microwaves to every child in school, but everyone knows that is silly. Just because there is, a strong correlation doesn’t mean that is the causation. I doubt very much that providing microwaves, toasters, fans etc. will increase the use of contraceptives. What is known is there is a relationship between the two and that is all a correlation tells anyone. Perhaps the number of appliances is reflective of socioeconomic status (SES) like education and occupation. SES is now generally accepted as a reflective measure of contraceptive use.  So while the study was looking for the highest relationship it turns out that the two strongest relationships were actually linked to a totally separate variable SES.

This example brings me to what is called the third variable problem in correlations. While there may be a strong relationship, it may be due to both of the variables being linked to a third variable. Contraceptive use and number of appliances in the house for example may have a strong correlation between the education, occupation and income levels of the house. This example is so obvious that many simply find it humorous to think about providing appliances as a way to increase contraceptive use.

Yet this happens all the time with problems that are extremely complex. Take for example global warming, not to get political, but if people place upon their heads the critical thinking caps they may find their maybe third variables involved and illusory correlations. It is possible that the earth’s orbit maybe increasing temperature, shifts in magnetic poles causing disturbances in the atmosphere are a few potential examples. It may be harder to find additional reasons for why the ozone layer is being depleted but what about certain rays and beams of gamma radiation blasts may potentially upset our environment. While these maybe less frequented views, it is still possible and I didn’t bother to dig anything on it to see if they have been disproven.

                Correlational data has always gotten the best of us, even as scientists. It is a starting point for some discoveries and the only means for scientific exploration in others. We simply cannot build two solar systems and have one with no vehicles and one with. 

 

Want to increase your energy? Get Sera now, they think critically.