3 Types of Causalities and Why They Matter

This is a guest post by Alex Pozin. He is a technology product marketer passionate about the potential of user-friendly and secure enterprise software to transform our productivity and work lives. MBA from UC Berkeley Haas School of Business. Tweets: @alexpozin

Regression_AnalysisTurn on your television to a talk show and you’ll see a debate that will often involve one side or another claiming that X causes Y. The situation is not much different in many businesses – a proponent of an idea or a solution will claim that their X will deliver lots of Y. In some cases, they may be partially right and still be generally wrong. How is that possible?

 Let’s answer by first exploring three types :

  1. Necessary: Let’s say you want to cook an omelette that your spouse will like. You will need a source of heat. We can therefore say that heat is necessary for making an omelette.
  2. Sufficient: Heat by itself is not sufficient for making an omelette. We need other ingredients. However, heat is sufficient to simply cook an egg. 
  3.  Contributory: If you use a high quality butter, a well seasoned pan, and fresh ingredients – all of these factors can contribute to your spouse’s enjoyment of the omelette. If you serve it when your spouse is hungry – he or she will probably enjoy it even more. Considering the audience of this blog – you’ve probably come up with many other contributory factors of your own.

 It’s not a surprise that debates can focus on contributory factors with intensity disproportionate to their significance (we’ll save the discussion of statistical tests of significance for another post). For example, were certain kids shows and movies really sufficient to cause Occupy Wall Street protests? Unlikely. Were these programs a contributing factor. I guess it’s possible. However, it’s reasonable to assume that not everyone who watched the programs protested (therefore not sufficient) and not everyone who protested watched these programs (therefore not necessary).

Perhaps there are other factors: the preceding financial crisis and associated scandals, rising income inequality, proliferation of tools to amplify messages and so on. That’s how one can be partially right and still generally wrong – by pointing to a cause that’s headline-friendly but logically insignificant.

 A contrasting and logical example comes from an excellent book on management of strategic programs by Terry Schmidt. Readers are encouraged to logically test whether the planned outputs are both necessary and sufficient to deliver on the intended purpose of the project. Isn’t this a clearer way to determine what’s important? Next time you’re debating whether to expand project scope, try out this test.

 Causality is a complex topic and we’ve barely scratched the surface. If you find yourself logically testing a dubious claim in the near future, I’ll consider this brief post sufficient.