Saturday, March 11, 2006

Decisions, decisions, decisions

I've thought and written about the decision-making process several times in this blog, starting when I discovered Gary Klein's recognition-primed decision model. Since then, I've seen opportunities to use both his approach and the more analytic Kepner-Tregoe approach I learned decades ago.

It would be simplistic to think these are the two choices, though. In Die Praxis des ganzheitlichen Problemlösens, Peter Gomez and Gilbert Probst make the point that there are three kinds of problems, each of which calls for its own approach. Simply put, I suspect they might use Klein's approach for what they call simple problems, Kepner-Tregoe for complicated ones, and forms of system dynamics for complex problems. (For the Kepner-Tregoe fans among you, Gomez and Probst define a problem such that it includes both problems and decisions in the Kepner-Tregoe sense.)

Even three are too few. I firmly believe in the importance of treating these approaches as tools, and we need sufficient tools to be able to address the problems we encounter. As with woodworking tools, it's not the quantity of tools that counts; it's how well we can use them to achieve the desired goals. A carpenter with 25 saws is probably not 25 times better at cutting wood than a carpenter with one, but that carpenter might find it useful to have more than one saw to manage different cutting tasks professionally and efficiently.

I found a new decision-making tool recently. While reading Geoff Coyle's A possible method for assessing the relative values of alternate system dynamic models, I discovered the Analytic Hierarchy Process, developed by Thomas Saaty. It appears to be a much more highly-developed, analytic approach to weighting inputs along multiple criteria than Kepner-Tregoe's simple approach.

While you can find professional software to lead you through the process and help you make the calculations, several of us on the J programming forum, including John Randall, Tarmo Veskioja, and Roger Hui, had fun developing a quick program to duplicate Coyle's results and possibly to use in future decision making. J fans might enjoy the dialog (look for most anything on LAPACK or principal eigenvectors) and the current version of the short script. Thanks to John, Tarmo, and Roger for helping make this work!



Post a Comment

<< Home