Correlation & Causality Explained

 

I learnt that the assumptions we make can turn out to be very destructive. Of course, wrong assumptions can occur due to many reasons, but in particular some are related to our inability to differentiate between correlation and causality. Correlation is a link between two measures, if one changes the other varies accordingly at the same (or the opposite direction). However, having two correlated measures does not always mean that one caused the other; actually both could be outcomes of a third variable.

Researchers usually identify causality by holding an experiment over two samples where one sample would be influenced by the factor under investigation and the other sample would be left untouched. This way the effect of the factor is identified in comparison to its absence. It is good to note that causality implies correlation but not the other way around.

Away from scientific jargon, my performance at work seemed to have improved recently. At the same time, a new challenge showed up at work which at the first glance might be the source of motivation that triggered improvement in my performance. After giving it a second thought, it looks like the slight improvement that I witnessed showed up as a result to submitting my dissertation, which took place at the same period and accordingly allowed me to dedicate more time and focus for my work.

The photo above is for the campus of Lancaster University, as I learnt the difference between correlation and causality while doing my MBA, this post goes to LUMS.

* Update Apr-2013: I forgot to mention that one of the ways to avoid building false assumptions is for us to simply ask questions and validate what we have in mind.