Social Science Statistics Blog

Amy,

I think "Occam's razor" and parsimony are overrated. For biology I can't say, but in social science, "complicated" is more believable than "simple," to me. If you can approximate reality with just a few parameters, fine. If you can use more parameters to fold in more information, that's even better.

See here and here for more.

On a separate point, as a political scientist I think it makes sense to understand the appeal of various beliefs such as creationism by considering the political and social forces supporting them. In this case you have a lot of churches and political figures too. It's not so much that someone's using Occam's razor or whatever to decide on a belief, as they're using these rules to justify a belief that it can be convenient to have.

Thanks for commenting!

Re the latter point, I definitely agree that there are many factors besides Occam's Razor that the layperson uses in this particular case, and I wouldn't even argue that factors like authority and emotional beliefs were less important than notions about the simplicity of an explanation. The question is rather about the subset of laypeople who have neither a pressing emotional reason to disbelieve in evolution (because they aren't religious, or at least not dogmatically so) nor a pressing political or social authority telling them they should (for the same reason). For those people, do their notions about simplicity of explanation play a role in their distrust of evolution as a viable explanation?

Your first point (and the links you gave) is quite interesting, and I definitely think that the question of to what extent "simple" explanations should be favored either by statisticians or social scientists is nontrivial. (As well as the debate about to what extent priors and hierarchical models impact "simplicity", but that's a debate for another day).

That said, the question of what social scientists and statisticians should do when making models is rather different than the question of what laypeople do do when creating models of the world, which is what I was more focused on. Whether it is "correct" in some objective sense or not, do people have and use Occam's-Razor like notions of simplicity? And if so, how and where? There is some research suggesting they do, but I'm not aware of any applied to their reasoning about complex scientific problems that they don't really understand, like evolution.

But I may have misunderstood your point. Another way to read it (sorry if I'm not getting it quite right) is that you might be pointing out that it is your personal experience to favor more complex explanations in social science, and therefore you doubt that people in general favor simple explanations. As they say, the plural of anecdote is not data, but introspection is a great place for a cognitive scientist to form hypotheses to test, so this point is quite interesting to me. :) To what extent do you think this intuition of yours resulted from your years of experience at working with statistical models in the social sciences? In other words, is it a byproduct of expert knowledge and training, or not?

Forgive me if this is too far off your post. There is, I think, another quite different way of approaching this question, which applies to many (but not all) who do not accept evolution. Nothing to do with Occam's Razor. Many Christians reject evolution in large part because many evolutionary scientists say that evolution proves there is no God. Such eminent scientists as Richard Dawkins and Edward O. Wilson come to mind. Christians, who admittedly are not experts in biochemistry, never mind quantum physics, say, "If that what these scientists say about evolution--and they're the experts, so they should know--then evolution must be false."

When leading evolutionary scientists say that evolution disproves God, what are faithful Christians supposed to think? If what I suggest is correct, then Profs Wilson, Dawkins, etc., are partly responsible for the widespread rejection of evolution. It seems to me that scientists who know the difference between science and metaphysics should publicly say something about this. Or maybe they have and the media haven't reported it--I don't know.

FWIW, I'm a Christian and a working statistician, and I accept evolutionary theory. The view I subscribe can be usefully summarised as evolutionary creationism. I was persuaded of this after reading Howard Van Till and Kenneth Miller, among others.

Scott

I was unaware that there are scientists that have publicly stated that evolutionary theory proves that there is no God - I'd like to see their statements before I feel up to commenting on that. Such a claim is, as I understand it, itself unscientific - there is no way that evolutionary theory could "prove" such a thing at all - hence my surprise.

But, that aside, I certainly do agree that many folks probably disbelieve in evolution for reasons completely unrelated to Occam's Razor. The question to my mind was whether that might play a role in addition to these other reasons.

In any case, thanks for commenting!

I wouldn't say Octam's razor is over-rated. I find advances in biology a lot more useful than Technological Determinism or Uncertainty Reduction Theory. Most Social Scientific theories could use some serious shaving.

I've been thinking about your blog for a while, Amy. Do you think it's fair to say that Octam's Razor would favor creationism if one takes the existence of God for granted? It seems a lot more reasonable to imagine an all-powerful being would simply fathom the world into being, rather than using evolutions intricate, round-about method. One would then have to explain why God didn't create the world he was after right off.