Thursday, May 21, 2009

Reverand Bayes

May the gods bless the good Reverend. As a practitioner and teacher of quantitative methods, I have been conflicted about Bayesian statistical methods. They have seemed really neato cool, but can I ask students to learn them on top of everything else? Should I invest the time?

I have found that Bayesian methods actually simplify my life, because model construction and end-user implementation for designs of any complexity are relatively transparent. In addition, they can represent nearly any ecological process. Sure, there is a bit of a learning curve, but much of that learning goes right to the core of statistical thinking and interpretation. I am currently using Bayesian methods to reinterpret an old data set, this time in a manner that mirrors the eco/evo interpretation so directly and precisely that I am still giddy.

I guess I will encourage students to consider these methods ... on a case by case basis.

1 comment:

  1. You never told me that you were proficient with Bayesian methods! I started up the slope but slipped a couple years back. Maybe it is time to try again...

    Your perpetual PhD student,
    Molly Mehling

    ReplyDelete