Reading Blink (2007-09-04/13)

Decisions are difficult. A good decision requires work; a bad decision brings consequences. More information sometimes helps, and--with the Internet--much information is available. However, consultant work gives me this perspective: more research delays a decision and increases its fee. Even if the decision is better as a result, the difference may be marginal and of less value than the lost time. During the delay no action is taken, which may have an opportunity cost. So I "sleep on it", stare out the window, or take a walk. I sometimes use these approaches--and my judgment--instead of additional information.

Malcolm Gladwell filled Blink: The Power of Thinking Without Thinking with anecdotes and studies about the effect of additional information on decisions. Some people assume that additional information always improves thinking; Gladwell cites situations in which that is not the case. An example is classical music auditions, which were biased against women until held behind a screen. The additional visual information actually distracted from evaluating the performance.

In the Afterward, Blink asks "When to Blink--And When to Think." The initial answer comes from Ap Dijksterhuis' studies on "unconscious thought" (which BBC News and New Scientist have summarized): think about simple decisions, sleep on complex ones. However, further reflection leads Gladwell to two qualifications: unconscious thought requires training, and statistical summaries suggest significant factors.

Gladwell gives examples of effect on judgment, both good and bad. Training improves judgment in the battle of Chancellorsville and the Millennium Challenge 2002. Biases have a bad effect on decisions we make in the "blink of an eye"; examples are the shooting of Amadou Diallo and the results of an Implicit Association Test.

In addition to training, statistical summaries also improve decisions by suggesting significant factors. The Wages of Wins suggests it is difficult to estimate summaries from incomplete observations of small differences. A report from Cook County Hospital exemplifies how statistics can isolate the factors relevant to a decision.

Statistical summaries also have their role in my consulting. I just completed a project requiring analysis of millions of records I imported from daily data and queried for mapping and evaluation. I use statistics, training, and unconscious thinking to improve the speed of my decision-making.

No comments: