The 'Man' Particle

So I’ve been working in Analytics for a while now…

Somebody once said that statistics is like a mini skirt; but don’t you so wish that it was almost as interesting???

And I’m told that statistics reveals as much as it hides; or maybe, it hides even more than it reveals – what do I know? – as a personal pet policy, I don’t do drugs, beef or stats.

A celebrated atheist is claimed to have said, “In god we trust, everyone else must bring data” – unfortunately there is no data to prove if he was celebrated, or an atheist or if he actually said these words. But nevertheless, my take on this subject is slightly different – yes I think that we must trust in God, but for everyone else, well…. keep reading.

Now in our Market Research and Analytics industry, I think there are two types of analysts – those with and without a stats degree; but it hardly matters as both are equally confused on the subject. I used to represent the comfortable majority who stayed as far away from stats as possible all this while but the annual goal setting carnival this year has forced me to bring stats in my life starting this year. So my goal for this year to learn stats and to “educate and enlighten my client of the benefits of stats in our engagement”. But the problem is – who’s going to educate and enlighten me of the benefits of stats? If any, that is. And I realize that my paralyzing fear of stats is not really helping; in fact, my fear of statistics always ranked high in my list of fears – way higher than my fear of heights and somewhere near my fear of clowns. And I’m also told that neither counseling, medication nor exorcism can cure one of one’s fear of statistics.

I think my first brush with Stats was in high school. I was then one of those kids with extremely low self esteem – I mean, I was so bad on self esteem that I didn’t even know how to spell it – but my enterprising dad made me believe that I was always ‘one in a million’ and hence, always different. And it seemed to work for a while till my stats teacher told me that even though I may be one in a million, statistically speaking, in India, there’s still a thousand more like me. And that pretty much plundered my self esteem for life.

And then there were those innumerable instances in my earlier companies where I needed to mentally convert a million into a billion (which I later found out was surprisingly easy) and to convert a billion into a crore (which was nothing short of mental hell!). Of course, we could all use our scribbling pads which would have greatly helped this conversion but we wanted to show the clients that we breathe and chew such basic mental calculations all the time. And so, our many mental maths used to turn clients’ billion dollar initiates into lakhpathi ones and vice versa, often in the same breath.

But with time, I started to like numbers and figures (the mathematical sort, you filthy mind J). I mean, data and data mining was kinda cool - in a way, data mining could be seen like finding a needle in a haystack or even a more complicated version of hide and seek. And making sense of data was also a very interesting exercise in creativity. And you always had an alibi if it falters - if all else fails, you could always blame it on incomplete data, the data source, the methodology or all of the above. But as imminent as old age, taxes and death, with great data comes great responsibility – and also not-so-great statistics.

One thing I never get about stats is the ‘confidence levels’ thing – I mean, historically, it’s always been a question of ‘to be or not to be’ - there’s just two arguments to any situation and none more. But with stats, you get to say ‘to be’, ‘not to be’ and also ‘could be’ – and that too, with various levels of confidence!! I mean, why are statisticians never fully confident? Do they also have self esteem issues??

But my biggest grouse against stats is something related to what I’d refer to as the ‘God Particle’. For the uninitiated, the ‘God Particle’ is that factor that’s responsible for the unexplained good things that happen in life – so you got a hike this year, the occurrence of which surprises you as much as it shocks your team members (or vice versa) – that’s due to the God Particle; if you win a lottery and you can’t remember even buying a ticket– that’s the God Particle at work; if you get an auto rickshaw willing to take you to office, charges you by the true meter and even gives you back change – that’s the God Particle on steroids.

So what does the God Particle have to do with data? Well nothing actually. But at the other end of the same spectrum is the Man Particle and that has everything to do with data. If the God particle represents all the best things that can happen in life, the Man particle represents the exact opposite - and the only thing suspect about data are the humans working on it. Data is actually contaminated all the time by the Man particle – and stats, which is defined as “whatever is being done to data - collection, organization, analysis, and interpretation of it”, in fact, recognizes the human factor in data; they even have a sophisticated word to call it – they call it ‘non sampling error’! But don’t let it fool you – it basically means ‘the unbelievable blunders that only we humans are capable of.

Few examples:

• For my summer internship, I was working with a research agency that was at the time working for a global leader in optical lenses. This client had just entered India and had commissioned our agency to do a market sizing study to find out the overall optical market in India, its different segments and the various players in each of the segments. After two months of data collection, the senior director of our agency was to present the agency’s findings to the India CEO of the optical company, who had specially flown down for this presentation. The director knew that findings were not too favorable for the optical company as the data showed that the company was always a distant No 3 or 4 in all the major segments in the optical market of India. To infuse a sense of cheer and to start on a bright note, the director identified a segment where the data showed the client company to be the market leader. The director cheerfully started the presentation by highlighting that “You should be pleased to know that our data says that your company is number 1 in the XYZ segment.” The room went quiet and the CEO blinked a few times amidst the pin drop silence in the room. The CEO then humbly told the director that they do not operate in that particular segment and never had any products there. Ever.

• Company A had to come up with a strategic response to a particular situation and was in a dilemma over deciding between two opposing courses of action. So it commissioned an agency to come up with the right decision. Now, one way of creating a strong impression on clients is to impress him with things that he does not know – so the researcher waxed eloquent about the power of statistics and did a hypothesis testing with the company’s data. Unfortunately, the researcher fumbled on setting the null and alternate hypotheses and unknowingly suggested the wrong course of action to the company. Later somebody else found the error but it was already too late as the company had invested heavily in the recommended action point. The irony however is not that the researcher got it completely wrong; the irony is that the company met great success by following that path.

• This happened to my friend’s market research agency. Both the top two competing firms in the ABC category were being serviced by two separate teams from this same agency. Both teams provide their respective clients with data and analyses on a periodic basis and this data is used by the client for strategy, competitor intelligence and study of market dynamics. Needless to say, contractual obligations did not permit these two agency teams to interact with each other in any ways whatsoever and hence were housed within the same agency on different floors. However, the two teams shared the same courier dispatch team of the agency. For the much awaited annual data analyses, the courier guy sent the dvd file of one company to its competitor and vice versa.

So as you can see, data in its purest form is like God – a source of unlimited knowledge and powerful information. But when humans intervene, well…..

So what do I do about data? Well, only in God, booze and money I trust.

For everything and everyone else, I keep a personal pet policy.

I don’t do drugs, beef or stats!

Amen.

P.S – This post is a work of fiction - I actually like beef.

-

P.P.S – By using the term ‘Man’, I refer to both genders – I mean no disrespect to the ladies; I know that whatever blunders men can make, women can probably do it better.

J