After doing a course in data science from Coursera, I was excited at the prospect of reading this book “Thinking with Data” by Max Shron, published by O’Reilly Media.
The book does start promisingly. The author argues the importance of considering the “why” before the “how” in projects involving data. Usually, with data in hand, teams tend to dive into the nittygritties of using the data than taking time to consider what question needs to be answered. The author proposes a framework for deciding on the right question to answer. With a convenient mnemonic CoNVO to outline the four areas that are in the scope of a project, the author explains the Context, Need, Vision and Outcome that should be defined before a project is started.
What had the promise of turning out to be a great lesson in the pre-analysis for data science, suddenly changes track into a philosophy lecture with the basics of argumentation from the next chapter onwards.
+ + +
- The emphasis on the “why” before the “how”. I totally agree with this, not only for data science projects, but for any act in life. It is true that one needs to validate the why before taking up huge projects and consuming resources for long time periods.
- The framework CoNVO that is defined is a sensible base to define for a data science project.
– – –
- With the good framework definition, the author could have made it more useful by giving lots of useful examples from all areas instead of a real-estate example specific to the US.
- Within the 93 page book, the author has spent too much of time (three chapters out of six) on the basics of good argumentation and reasoning. This leaves us with hardly anything to benefit from the book.