Use Data to Combat Natural Human Biases

Maria Popova is a Bulgarian writer, blogger, and critic known for her blog BrainPickings.org which features her writing on culture, books, and eclectic subjects off and on the Internet. According to her article, "How Our Delusions Keep Us Sane: The Psychology of Our Essential Self-Enhancement Bias," there sure are a lot of biases that make up our thought processes:

• Self-enhancement bias

  This is our systematic tendency to forgo rational evaluation of our own merits and abilities in favor of unrealistic attitudes that keep our ego properly inflated as to avoid sinking into the depths of despair.

• Superiority bias

  This is our tendency to judge ourselves less harshly than we do others and to see ourselves as unique, special individuals amid a homogenous, dull crowd.

• Illusion of control

  This is our hindsight's inclination to attribute our successes to ability and our failures to luck.

• Optimism bias

  This is a mental construct that deceives us into believing everything will work out in the end even against all odds and statistical data to the contrary. The bias, however, disappears when we observe others.

• Confirmation bias

  This leads us to notice more of the information which confirms our beliefs and less of that which contradicts them.

• Hindsight bias

  This causes us to retroactively revise our own predictions in the face of new information and claim that we always saw it coming.

• Self-serving bias

  This lets us take credit for all the good stuff that happens to us but blame the bad on external circumstances or other people.

• Self-enhancement bias

  This works just as the name suggests — it enhances your view of your self.

I believe it is in this spirit that visiting research fellow, Mickey McManus, wrote "Simulating the Unexpectable (or Lie to Me)." Mickey is making the point that, like biases, should software help build our confidence? Should software complement our natural human tendencies? My answer is no. As a scientist, I reject this notion. The bias in our thought processes underscore the importance of data. In his keynote address at the opening session of Autodesk University 2012 Autodesk CTO, Jeff Kowalski, quoted famous engineer/statistician, Edward Deming, as saying "In God we trust — all others bring data." The data provided by visualization, analysis, and simulation tools should be real, not inspirational. Data that is not real is not science. That type of data is only for motivational speakers who live in vans down by the river.

The search for truth is alive in the lab.