It’s not the model in error: it’s the data!

You’d think an engineering association publication would be a bit less obvious.

So the problem, in a nutshell, is that not seeing is not believing. Since the deaths can’t be identified or even measured, the Times’s editorialists are treating them as if they’re not occurring. They really ought to know better.

You cannot see it, you cannot measure it, so you’re stupid not to believe it anyway? It is just like in climate. Data means nothing if you have something better.

So they ought to appreciate that if dose-response models predict a certain level of fatalities, those fatalities must be assumed to be occurring, even if they aren’t seen.

It’s not that the model is wrong, you see, it’s the data that’s wrong. The data must be forced to fit the model whether it wants to or not.

In this case, Bill Sweet is taking up the anti-nuclear religion. He ends his screed with the zero risk assertion – another engineering absurdity.

You really have to wonder what would happen if engineers allowed a model to over-ride actual experience and data and expected to be able to craft zero risk devices.

Comments are closed.