Rooting out cognitive biases in the intelligence analysis business needs much more than sticking a post-it note on your computer screen. One of the more interesting aspects of the work IMSL does is helping organisations develop management techniques for the intelligence process. With one or two notable exceptions, most intelligence analysts and most intelligence organisations are very prone to cognitive biases. There are many unexploited opportunities to construct methodologies to avoid cognitive biases. These can be procedural or they can be prompted by the software tools we need, or they can be implemented by managers in a variety of forms. Cognitive biases affect the decision making process, and the prior to that, the analytical process itself. In intelligence analysis, cognitive bias can result in life or death. It’s damned important. One tiny slip can waste money, time and real peoples’ lives
So in general terms I’m always seeking techniques that intelligence managers can apply to themselves , their staff and the processes that occurs in intelligence organisations. Sometimes my own cognitive bias needs overcoming too… One of my biases is against the HR policies of large corporations which too often I’ve seen be pretty bloody awful. So at first I was resistant to reading a book by Lazlo Bock, called Work Rules, the HR lead at Google. The perception of Google’s infamous hiring techniques and the silly questions they were alleged to ask potential recruits put me off. I’m also biased against trendy books that suggest ways to manage, period. Now this book addresses the broader challenges of managing people, but there are some really interesting approaches to general decision making processes that certainly have a very valid application to the intelligence analysis process.
I think companies that produce intelligence analysis software can do much more to construct in-built mechanisms to reduce cognitive bias, and I’ll write about this more in future posts. The truth is that intelligence managers need all the help they can get and the pretty wonderful software tools available to help undertake analysis and provide analytics can do a much better job of nudging analysts and their managers away from cognitive biases.
The book also underlines the huge importance of understanding statistics as a skill that can be applied by analysts to avoid bias. I’m pretty horrified at the poor statistical understanding of many intelligence analysts I meet, and its absence in formal intelligence analysis training. This too could be addressed better in commonly used analysis tools.