Insights: Statistics

Presenting Data – Fighting the Russians

IMSL is currently spending a lot of effort developing the mechanisms by which we present analysis of data to our clients.   I think everyone agrees that the same-old-same-old Powerpoint is too crude, too boring and too limited in its presentation tools.  Over the next few weeks we are going to show you some of…

Rubber Intelligence and “tiebrapolnud”

This is another post in the series looking at statistical analysis and its surprisingly useful role in intelligence. You might find statistics unattractive as an area of study – I hope to change your mind if you are an intelligence analyst.  Once again I’m looking at WW2 historical experience, because the innovative processes utilised have…

Maths and Statistics of Spies and Terrorism

Good  item from “More or Less” a year ago, but just a relevant today.  It’s worth a listen as it discusses the weaknesses of surveillance operations, due to the “rareness” of terrorists and the implications to surveillance legislation.

Correlation and Causation

As I’ve blogged about before, an understanding of statistics, I think, is vital for an intelligence analyst. The “German Tank Problem” which I discussed here and here is simply one example from WW2. Human nature (cognitive bias) often gets in the way of understanding statistics. There was a great piece here about a coroner linking…

Intelligence Use of Aspects of the Scientific Method

I’ve blogged before, recently even, about the challenges currently facing parts of the scientific community with regard to the application of the scientific method, and the  chagrin I feel when compared to aspects of the intelligence community which could learn a lot from adopting aspects of a formal scientific method.  When did you last read…

The German Tank Problem and Intelligence, Part 2

I’ve been scratching away at issues surrounding the interface between “intelligence analysis” and “statistical analysis” since my first post on the German Tank problem here.   To be clear, I’m approaching this from an intelligence analysis perspective not as a statistician. So I have about 5 or 6 posts to make on this, which I’ll…

Intelligence, Statistics and the German Tank

A good intelligence analyst needs a golf bag of skill sets. One of the slightly rarer ones is statistical analysis.  Quite often the data collected with regard to intelligence is incomplete. Statistics can be used to extrapolate from that incomplete data to come up with something useful. Statisticians are a funny breed and this overlap…