While discussing the blog with friends, there was a fair bit of confusion about what the word really meant. Some thought that I meant realtor and then I realised that many others would probably have the same question, hence this piece. Relator is the Latin word for storyteller. Storytelling is quite the rage right now but in my mind, almost all data has a story that is hiding behind it and it is our job as analysts to uncover and weave those pieces together to create a cohesive output. It doesn’t really matter what tool or technique we employ but getting to the bottom of the story is what is most important. One of the best known analytical minds of our time, Nate Silver believes numbers cannot speak and it is upto us through our analyses to “imbue” them with meaning.
You may have questions about the applicability of this style to different contexts but in my view, being able to analyse data and generating output that is comprehensive and coherent to the end-user is one of the most important aspects for data scientists / analysts to consider. One of the earliest books on the strategic aspects of building a cutting-edge data science/ analytics programme was "Competing on Analytics" and the section on skills listed good communication among others. Though this book was first published in March 2007, I don’t think this fundamental requirement has changed. Many of us are very skilled on the quant side of things and coding but what we need to have in addition are excellent communication skills and being able to simplify the data science lingo so that the analysis is accessible to all. We assume at times, that this skill is just acquired on the job but my experience has been that we need to be intentional about developing those skills and the best part is that it would help us show off some of those amazing analytical and coding skills to the best possible advantage.
Data Science has taken the world by storm and some people are already in the thick of things while others are just curious about it and all the terms that are bandied about. The well known computer scientist Judea Pearl who wrote a book called “The Book of Why” refers to the discipline as the “Science of Causal Inference”. “The new science has spawned a simple mathematical language to articulate causal relationships that we know as well as those we wish to find out about. The ability to express information in mathematical form has unleashed a wealth of powerful and principled methods for combining our knowledge with data and answering causal questions.” Whether it is “simple” or “principled” can be a matter of opinion but this is one of the many ways to think about this simple and principled science.
This blog I hope will provide some insight on the strategic potential of data science as well as delving into some those tips and tricks that I have used over the last fifteen years in my work. In my first job as an analyst, I quickly learned from people who had been in the industry longer than I, that this is a science for sure but getting the results to align and make sense to the end customer often required a bit of art as well. It is not quite the same as pulling a rabbit out of a hat but as you grow in knowledge and confidence one does become a better judge of how the science can be applied to satisfy the scientific requirements while being useful to people who will use it or generate insights that will be useful. I am not quite sure if the analogy is perfect but I always think that most disciplines have elements of both the sciences and the arts. One of my favourite books of all time is a book called “The Agony and The Ecstasy” by art historian Irving Stone which is a biography of the great Renaissance artist Michaelangelo. As I read the book, I realised that Michaelangelo who was a consummate artist, still needed or wanted to understand more about the structure of the human body to be able to hone his sculpting skills and he indulged in vivisection to understand more about muscles and so on. Much like the artist Michaelangelo, we will need to employ the right blend of science and art to be able to generate usable analyses and recommendations.
Running analyses to determine which analytical framework is best suited to the data and the problem being studied is something that very often subjective and some experts like Scott E. Page advocate a many model approach while analysing. I am not opposed to this approach and with computing power that we have access to, ensemble techniques that combine frameworks are increasingly used. Otherwise it may be that the problem that you are working on necessitates a particular framework. We will delve into some of those questions as we go along in this blog.
I have always enjoyed solving problems and now as a professional, it is immensely gratifying to be able to employ some of those skills to get to the bottom of a challenge that your client is facing. Once you have done an amazing job analysing the data, you owe it to yourself to communicate all you have done succinctly, using minimal jargon to maximum benefit. So I would greatly appreciate you joining me, as I embark on this journey of exploring data science, demystifying it, and exploring the strategic power of storytelling whatever the style or the purpose of your analysis. Feel free to comment below and let me know if there is something you would like me to cover in particular.