Every now and then some readers write to me questioning the concepts of decision theory, probability, and statistics. A few of these are genuine questions about the underlying concepts, but most have an agenda. Hidden agendas come into play when the results of formal theories conflict with some deeply held belief. That’s okay because I firmly believe that a person’s faith is made stronger and better when it is consistent with physical observations. Believing in something that is nonsense just to show how much faith you have can be a fun sport, but I have other things to do. Not everyone shares my orientation and that can lead to misunderstandings.
Many of these misunderstandings can be avoided by being careful to remember the area of expertise being discussed. To take a provocative example (and one that has filled my inbox), neither science nor decision theory can directly decide if sexual abstinence outside of sanctified marriage between a man and a woman is morally preferable. That is a matter of faith. However, when people who accept that abstinence is a good thing try to sell the concept as a better method of birth control, we take the gloves off. As a method of birth control, it is an abysmal failure.
“What?” the proponent cries, “If you don’t do it, you don’t get pregnant. It’s that simple.” Well, no it’s not. Using that pseudo-logic, I can say that the pill never fails and condoms are 100% effective. The point is that once you start to make statements that have measurable consequences, then the powerful tools society has developed over the years can be applied to determine the likelihood that the statements are correct. Abstinence proposed as a moral value is one thing; abstinence proposed as a prophylactic measure against STDs and unwanted pregnancies is quite another. The first is a statement of values; the second is a testable hypothesis.
That distinction is important to keep in mind when considering sensitive issues, particularly issues which you personally feel are beyond question. What are your blind spots?
Ironically, the development of decision theory is attributed to the Rev. Thomas Bayes, who was a sincere man of the cloth. So people who maintain that only secular humanists and atheists rely on decision theory are wrong from the get-go. (I say attributed to Bayes because many other people also contributed, including the great Laplace, and Bayes himself might not have agreed entirely with the concept we now call Bayesian decision theory since he really didn’t present it the same way we usually do now.) In accordance with English practice of the time, the good Reverend was buried in a place reserved for non-conformists – another interesting irony.
Please keep in mind that this column deals with things in the real world that can be measured. I present no opinion of the existence or non-existence of a soul, a divine being, or any other unfalsifiable postulate. So when President Bush urges that conjectures such as the latest version of creationism be taught as a theory, I can dismiss his mistake as his personal failure to understand the difference between a falsifiable theory in good standing and an unsupported conjecture. This is a different type of misunderstanding than wrongly predicting the presence of WMDs in a foreign country. I can attack the first type of misunderstanding by emphasizing that we need to agree on definitions, but frankly, I’m pessimistic about making much headway. Politicians of all stripes are willing to bend the language to fit their needs. They always have been, and I see no reason to expect them to change.
The second type of error, the misuse of data, is where I concentrate the majority of this series. By alternating between fun puzzles that are chosen to illustrate various principles and drier columns (such as this one) that look more at the theoretical aspects, I hope to help at least a few people pick up the tools that will help them evaluate real issues in their real lives. To emphasize that these techniques are not just academic exercises about hypothetical horse races or phony diseases, I try to use emotionally laden examples where logical decisions bump up against ingrained predispositions.
The use of inflammatory examples does not represent a hidden agenda on my part. The only agenda is to provide food for thought.
For those who wish to delve further into decision theory without wading through a lot of equations, I have posted a tutorial on elementary decision theory. It shows examples of faulty physicians’ diagnoses (important for those considering surgery) and how to evaluate anti-terrorist activities (important for everyone). That tutorial can be found here.