Building a Rational Decision Making Process

Overview


Given that rational decision making can be a highly theoretical subject, it is hardly surprising when decisions are made that fall short of the ideal standard. Knowing this, many organizations explicitly create a decision making process in order to move their decisions closer to the ideal.

Most decision processes are designed specifically to avoid the known behavioural traps.

Example Considerations


There is no single best decision making process. However, a number of possible recomendations have emerged as

  • Avoid Decision Paralysis : knowing that having too many decisions can lead to decision paralysis means that any decision process should strive to limit the number of choices when making a final choice. A sample best practice in this regard is the mece framework.
  • Utilize the Wisdom of Crowds : leaders should build a process that gathers inputs from many diverse viewpoints, while striving to avoid the traps that occur from social proof. (For a mathematical treatment of wisdom of crowds, please see ensemble modeling)
  • Prepare to be Wrong : recognizing that decision making occurs within an inherently probabilistic situation, decision makers need to be cognizant of any assumptions that drive their decision making, and to test and contiually monitor whether any new information has invalidated those assumptions. As time progresses, it may be necessary to make tactical adjustments to the decision, or possibly to abandon a decision entirely.

    Bias to avoid:

    • Overconfidence
    • Sunk Cost Fallacy
    • Regret Avoidance
    • Confirmation Bias
    • Greed and Fear

  • Provide Appropriate Challenge : all decisions must face challenge at some point. If during a decision process, a particular decision sails through the process without any significant challenge, the team should consider assigning someone to challenge the position.

Use of Models


Using models during a decision process can often times be helpful. In addition to providing insight into the situation at hand, models can be used to short-circuit some of the behavioural traps that a decision can fall prey to.

However, models come with their own set of risks. ( see model risk ).