> Code review is not the way to validate statistical inferences.

But it is a way to invalidate the inference, which is what the defense is interested in.