Common Tricks Data can Play on You
An Illustrated Collection of 15 Statistical Fallacies to Watch Out For
We all use data almost daily, be it digesting it in a news article or sharing it when deciding next steps in a work project. The problem is that our reasoning is filled with cognitive mistakes and biases, as renowned psychologist and winner of the Nobel Prize in Economics Daniel Kahneman has shown. Sometimes it’s accidental and sometimes, in our post-truth world, data is deliberately distorted to serve specific agendas.
Statistical fallacies might sound quite niche and not like something that would affect the general population, but these mistakes, deliberate or not, are happening all the time in areas like medicine and politics. A recent study in Psychological Medicine showed how several biases were causing a compound effect on the outcome in 105 different research studies into antidepressants. One of the fallacies was publication bias: 98% of the positive trials got published but only 48% of the negative ones. This kind of bias could in the long run conclude that a treatment works when in fact there are cases where it doesn’t, as a BMJ modeling study showed.
With this in mind, we identified some common ways data can be misconstrued and designed a poster of 15 data fallacies to watch out for.
We hope people can use this advice to draw better conclusions in their work and everyday lives. You can read more about these statistical fallacies here and also download the poster for free.
