Data Analysis and Stats

A common theme across MBA programs is exposure to and at least basic skills in, statistical analysis. The value of this can’t be understated, if you’re already working as a quant or did significant amounts of stats in your undergrad you’re not going to be exposed to many new concepts here but for those that haven’t done a lot, this can be a real eye opener. I noticed it particularly with friends that came from the more qualitative industries, marketing or legal, they all highlight that this is where their industries are headed. Gone are the days when a marketer could say “give me a product and I’ll sell it to everyone”, the industry is driven by analysis: what defines a segment in your industry; what is their willingness to pay; how price price sensitive are they? No one is going to invest in your campaign unless you can demonstrate what you expect to achieve and how you are going to measure the results and the answer to both of those needs to be statistically rigorous.

At MBS I did two primarily statistics based subjects (a core data analysis subject and an elective on market research) which, content-wise, can effectively both be split into two parts. Firstly you have the number crunching component: how do you run a regression, how do you test for correlation? The second (and in my opinion far more relevant component for managers) part is what questions do you ask and what you do with the results? Understanding the mechanics behind how data analysis works is handy but lets face it, most people don’t enjoy crunching numbers. I’m convinced there’s something behavioral in there that links back to primary and secondary school and the finality of either getting the right or the wrong answer that pushes people away. I see why there needs to be a component of this in the course but even as an engineer, I’m rarely going to do this kind of analysis myself, I’m going to hire someone who does it professionally and let them do the math. Where an MBA grad can add value is in the interpretation.

I’ve learned that interpretation of statistics is a dark art, far more qualitative than you would expect and, like anything where subjectivity is involved, this sets up a system that can be manipulated by those with the will to do so. Nowhere is this more evident than in media polls; read through a newspaper and look for the headline that screams “Labor drops 2% in two party preferred rankings” then look for the fine print. What was the sample size, how was selection controlled for uneven distribution of voting preferences, what’s the standard error? Unless it’s backed up by rigorous analysis, a headline like that is at best crap and at worst outright misleading. Print media in Australia still plays a major role in shaping public opinion in Australia but its that kind of reader manipulation that destroys the trust of the public when the realize that they’ve been had, the problem is, without an understanding of how statistics works, people miss the deception.

I was talking about this with a friend and he pointed me to a TED talk by Arthur Benjamin where he suggests that the problem is that we’re teaching maths wrong. He argues that the progression we have in schools at the moment (arithmetic -> algebra -> calculus) is going in the wrong direction. Calculus is great, as an engineer or a physicist it’s fundamental to our professions but for 90% of the population, it’s completely irrelevant but statistics (and the associated field of probability) is something that we use every day of our lives. I won’t steal his thunder, it’s only 3 minutes long and like most TED presentations, it’ll leave you thinking.

But back to the value that a stats course adds to an MBA… One of the things that an MBA does really nicely is combine the general knowledge and some big picture perspectives with the specialist insights that you have from your experience prior to the MBA. The ability to look at the micro and macro levels of things is one of the core skills that an MBA pushes you to develop and statistics knowledge helps you ask the right question, get more valuable insights when you interpret information and understand the difference between opinion and something that’s statistically significant. You may not have the best maths skills in the class but that doesn’t matter, being able to understand how the information is relevant and how you can use that information is what is important. Some schools understand this difference and finding one that does will help you have more than a primary/secondary, right/wrong  experience. Looking back, it’s one of the key things I would look for (content wise) in an MBA program.