Naked Statistics: Stripping the Dread from the Data
I bought this book in an airport to bridge a 4-hour delay for my next flight. Going by the title, it does not sound like the best book to read in an airport. The context was that I was returning from a data science workshop; the theme simply said ‘Sampling’. The knowledgeable speakers often referred to the ‘Central Limit Theorem’ (CLT) and the importance of representative samples. When done efficiently, inferences from a small number will provide near accurate insights about the larger population. I quickly browsed through this chapter in the book and it became immediately clear, why CLT is the foundation for sampling. The chapter was in plain English.
I grew up in the days where physics was an enchanting subject, smarter ones excelled in mathematics. The horizon held two career options; engineering science and medicine. George Gamow and Isaac Asimov adorned the bookshelves. Statistics existed somewhere in the recess of all things bright and beautiful at a considerable distance; a pariah if you will. The meaning of inertia was intuitive; however picking a red coloured pen from a bag that has red, green and black coloured pens did not make much sense. Putting it back into the bag and picking it again for the second time was absolute nonsense.
The averages was ok, the frequency distribution was monotonous and standard deviation, a mere formula. The bell curve made little sense but was a dread that levelled the horses and donkeys to mules. The person who wanted to pretend studying science majored in statistics.
We now live in an age where statistics is like internet. Bayes is the new poster boy, people trip over each other to estimate the probability of picking a red pen over a black one alluded to earlier. Suddenly, in this brave new world if one does not fully understand averages, standard deviations or a normal curve, there is a high probability of missing the bus and a fair chance of being stranded. Who is the villain who has poured elixir into statistics much to the wrath of loyal pure science and engineering populace?
Big Data comes up as the foremost candidate amongst the suspects. The data that the world is drowning in and the data scientists are bending backwards to make sense from it as it accumulates by the millisecond. Mathematical models it is said, is the way to do it. Statistics lends itself as a handy tool for the algorithms, pliable enough to draw meaningful conclusions. Even beyond ‘Big Data’ statistics is the new mantra all of us need to know to understand why by surveying 100,000 voters; the psephologists predict the winner with remarkable accuracy or for that matter, how a bank determines and detects the customer attrition percentages.
Naked Statistics: Stripping the Dread from the Data by Charles Wheelan is what the doctored ordered for the non-professional to understand the world shaped by BIG Data. The author’s candid admission of dread for numbers attracts the reader to hold onto the book and shed any fearful thoughts of seeing a sigma or a theta with funny equations to make it ‘clear with an example’. The truth is such clarity makes me look pie faced. Wheelan is smart to know this sinking feeling. An autobiographical quote says all “I am a senior lecturer and policy fellow at the Rockefeller Center at Dartmouth College, a former correspondent for The Economist, and the author of assorted books that attempt to make serious topics more accessible (and even fun).” Statistical analysis is much an intuition as Newton’s second law of motion.
Many otherwise baffling events that appear to be con games (to us) become clearer when you look at them using statistics. The famous ‘Monty Hall Problem’ from the show ‘Let’s make a deal’ does improve one’s chances to see the car than the goat behind the door, the second time over. ‘The Central Limit Theorem on which Statistics stands firm on, is given an insight that one would not have seen before. The principles of probability, inferences, hypothesis testing is stripped off the technicalities; the amazing choice of examples (Use Cases to pun) makes it delightful to read and quote as well. I am tempted to quote from the book, however I will resist it for now to allow you to enjoy the world of Statistics. You may even appreciate why Wheelan has used ‘Naked’ in the title.
This book is a ‘must read’ and I strongly recommend it. Wheelan does a better job than what most college professors can in a classroom. A ‘whodunit’; you will read from cover to cover in one sitting. I will not be surprised if this book shows up at the top of your bookshelf or on your Kindle shortly.
Reviewed by :
Vishwanath Thanalapatti, PMP
Vishwanath is a banking professional with global experience. His focus in recent years has been in corporate banking and analytics related to corporate banking. His vast experience includes capital markets and risk management. He reads extensively on subjects as diverse as philosophy, blockchain technologies, humour etc. He keeps mentally and physically fit by practicing yoga and running.