Causality vs. Co-relation

Anybody who has dreaded algebra and Calculus like I did throughout my years at School and then at Grad School will perhaps agree that a lot of the fear comes from a lack of understanding. Teachers expect students to learn formulas, difficult concepts and apply them to problems printed in text books as part of a syllabus. But how many teachers really take the time to tell you "why" this is relevant for you? Very few. Making it hard for people to acquire a skill that I have come to realise is critical for professional success.

I would have gone through life with a tremendous handicap but I had a bit of last minute luck. During my post-grad, a mathematics teacher decided to finally educate me on the reasons behind mathematics. and once again yesterday when I was reading "Freakonomics" by Levitt and Dubner I learnt some more- about co-relations and causalities. On page 163 of my edition the authors ask "Let’s say that you want to ask the ECLS data a fundamental question about parenting and education: does having a lot of books in your home lead your child to do well in school? Regression analysis can't quite answer that question, but it can answer a subtly different one: does a child with a lot of books in his home tend to do better than a child with no books? The difference between the first and second questions is the difference between causality (question 1) and co-relation (question 2). A regression analysis can demonstrate a co-relation but it doesn't prove cause" In other words there maybe a co-relation between books and a child's grade but causality alone can give that co-relation a meaning beyond just statistics. That is why the authors go on to add that “A regression analysis alone can't tell you whether it snows because it snows because it is cold, whether it’s cold because it snows, or if the two just happen to go together….. What we really want to do is measure two children who are alike in every way except one- in this case, the number of books in his home- and see if that one factor makes a difference in his school performance. “

This is what differentiates a pure statistician from a person who has the skill to look at numbers in a meaningful way. Unfortunately till today this skill is a rare one. And it became even more evident when I was reading an article published in “The Mint” a couple of days ago. It is titled “Wealth effects of the phone ring begin to widen”. The article – covering a good page of the newspaper spends ninety percent of the space throwing out statistics with no effort to establish a causality that would tell us why the cell phones are having a positive effect on smaller town economies. It just makes statements like “Indian states with higher mobile penetration can e expected to grow faster with a growth rate 1.2 percentage) points higher for every 10% increase in the mobile penetration rate” It quotes the MD of India’s largest cellular company saying that 2% of the country’s growth GDP growth has been contributed by the growing telecom sector. The question that needs answering here is – which 2%? Education, rural commerce and consumption, employee productivity? If we had the answer to that question then we would know where to push the mobile penetration and how. If the correlation could be extended to meaningfully establish causality then we could make the growth even faster and more efficient.

After it has quoted every industry sponsored study and the head of every organization with a stake in the growth of the mobile business , does the newspaper get down to giving some explanation of why the mobile phones maybe considered as drivers if growth. “Mobile phones can perform in under served areas and regions in the same as fixed lines did in many developed countries…widen markets , create better information flows”. But the article would have been much more meaningful if it had quantified just how much more this information spread can be.

But then maybe I am expecting too much from a newspaper article. It’s job is to give you an overall picture. They can be absolved of this responsibility but the teachers in our class rooms should never be. They should equip students so that they know when to see statistics for what they are, question them if needed and apply them where most beneficial. Because if they can not do that then it will always be lies, damn lies and statistics.