Ockham's Razor

[wazz]

I have no understanding of full and reduced models in statistics, but what you've said sounds both right and wrong, if you see what I mean - right in that it sounds analagous, wrong in that the extra information is making the proposition likelier rather than less likely.