Monday, November 21, 2011

Decision-Making, Limits and Pitfalls

Following on from The illusion of validity.

Kahneman's discussion leads to the modification of this decision-making graphic that I created identifying the four different modes of decision-making based on their present:futurity, tactical:strategic, incremental:consequential and reversible:irreversible orientations. All decisions are made in a context of more or less time available and we tend to default to that mode of decisioning which is most relevant to the time constraint rather than the mode of decision making relevant to the nature and impact of the decision as we ought to do.
This new version incorporates the delimiters associated with each mode of decision-making. Experiential, for example, depends on creating useful narrative stereotypes but there is a boundary of stereotypes beyond which we will not accept for any of a variety of reasons - lack of utility, conflict with values, incongruence with assumptions, etc. The other modification is at the center of the graphic. Once we have digested inchoate data and made sense of it, i.e. arrived at a decision using the four modes of decision-making, we then undertake action based on that decision. There is the direct set of actions necessary for the here and now, and then there are the more policy related actions relevant to the strategic, future state.
A final iteration prompted by Kahneman's comments. Each mode of decision making (heuristic, experiential, analytic and profound) is subject to biases and effects pertinent to that particular mode. In Kahneman's article he is focusing on the tendency to lock on to narrative coherence. We like stories that make sense and hang together and we are biased towards those coherent stories regardless of what else we might know about the utility of those narratives (stereotypes). So what are some of the common biases and effects?

4 comments:

  1. Hello - I would like your permission to use these diagrams for a presentation I am making regarding medical decision making. It is a presentation from one physician to other physicians regarding how to choose particular products during procedures. Please let me know if this is ok and thank you very much.

    ReplyDelete
    Replies
    1. Yes, you are welcome to use these diagrams. Please reference Charles Bayless or Decision Clarity Consulting as the originator.

      Delete
  2. I understand the concept of how time fits in this model but can you clarify why it is represented the way that it is? Is is it showing you have more time at that top? Is it suppose to represent a scale of time?

    ReplyDelete
    Replies
    1. To be honest, it is really just a reminder that Time is an important variable that has different affects depending on the context. I put it in the top right to mark time passage with regard to both acquisition and execution.

      In other words, it generally takes longer to form and act on profound beliefs than it does to acquire useful heuristics. A child of six already has a huge portfolio of heuristics, a reasonable portfolio of stories (experiential), some exposure to structured analytical thinking and usually only a raw concept of profound values. As you age all quadrants keep getting further populated and refined but the Profound quadrant probably takes the longest to mature and also takes the longest to refine.

      There is also an implication about the speed of execution. Heuristics are acted on virtually instantaneously and without much conscious thought. Experiential requires some cognitive matching effort (what pattern of events have I experienced in the past that match this current pattern and how do they differ?). Analytic is even more open ended (what are my goals, what are the facts, how do they relate to one another, what are the actions that can change the current facts to achieve my stated goal, etc.) Profound takes even longer because you are typically working with multiple goals, lots of uncertainty, different trade-offs, etc. You can make snap judgments on a profound basis, but usually want time.

      Delete