I’ve been lucky to be able to snag a week off for some reading and reflection.
On the reading list – “Canada and the United States: differences that count” – which came on the recommendation of Joseph Carrabis. It’s a very good book.
I trolled “Competing on Analytics” and “Supercrunchers” for golden lines that might resonate with non-scientists.
And, on the back end, I want to finish “The Volatility Surface” and “Computability, Complexity and Constructivity in Economic Analysis”.
I’m happy to read that others are having the same problems I am. “False Induction” is one, especially when you’re dealing with volatile data like web analytics. How do you quantify the rate of uncertainty once you’ve observed a pattern? Can you predict the degredation of an observed pattern if you don’t fully understand the cause?
Is it even computable?
Back to work Monday, I’m afraid. 🙂