Lies, Damned Lies & Statistics

Making sense of research data

I’ve never been a huge fan of numbers in my work, even though I love math, and one of my favorite books of all time is about Ramanujan, perhaps the most brilliant pure theoretical mathematician since Newton.

It probably starts with my basic personality type. I’m an INFP in the Myers-Briggs Indicator, and more drawn to the human, intuitive, feeling parts of my work than the number-crunching, data analysis parts.

Having that self-knowledge, I’ve acquired lots of great partners over the years to help me with the stuff I don’t enjoy doing. Invariably, they’re better than I am at what they do and they bring a different perspective to every discussion.

As PPM analysis has inevitably begun to affect all aspects of programming (except commercial load — what a shock!) and as perceptual research has been an expendable budget item, the value I place on these partnerships has been enhanced.

First, because not everyone who analyzes data is correct. They’re not all equally insightful and skilled in their analysis.

Second, because of each of the points raised in this article, How To Lie With Statistics.

It is definitely worth two minutes of your time.

I would love to exchange resources with you. Who do you absolutely trust as your research and analysis partner?

It may be the most important working relationship you consider today.