Good and bad science

If marketing is a science, we consultants owe it to our clients to be scientific. But pressures of time and especially the attraction of IT applications that do our jobs for us (and which we don‘t fully understand) mean that we sometimes lose sight of how we arrive at our ‘scientific’ conclusions and we leave ourselves open to criticism.

How often do we hear misleading statistics bandied around in the media, or see portions of graphs which have been cunningly selected to prove the presenter’s point? The continuing row between the UK Government and the Office for National Statistics and the propensity of even the most respected broadcasting organisations to dupe us with colourful histograms and pie charts are just two examples.

I was reminded of this by an interesting clip that has been doing the rounds over the last few days. It’s a presentation on the growth of Walmart stores in the US built in a great application called Modest Maps. Notice one thing at the end, though. It is the presenter’s intention to show how the brand ‘spreads like wildfire’ and therefore the rate of new store openings is what’s critical. The time bar seems to run in a linear way (1 second = 1 year) until it gets to 2006 which lasts a lot longer. As a result the presentation is less impressive than it might have been with a final burst of frenetic activity. Still a pretty interesting free application, though.

Ben Goldacre writes untiringly on Bad Science here http://www.badscience.net/ and in The Guardian.