Originally posted by _V_
View Post
Like many scientists he seeks information to support a hypothesis he has already emotionally invested in.
a/Look at the question, work out the confounding factors, design a robust instrument of collection, recruit a population large enough to give you a reasonable estimate of the effect with adequate precision, adjust for confounding factors, model the data, interpret the results.
b/ or cobble together a poorly designed, underpowered 'study' that gives you the answer you want.
something tells me he went for b/
Comment