A cracker company isn't pleased when it finds out the results of a survey it has taken. According to the survey, although customers would rather have a cracker than nothing at all, those same customers would prefer peanuts to anything else!
A junior employee decides this is his big chance for promotion. He claims to his boss what the survey really saids was that customers prefer crackers to peanuts. How in the world could he come to that conclusion?
The customers want peanuts above all else, but if they can't have peanuts, they can have either crackers or nothing at all.
ReplyDeleteIs it because that last sentence in the first paragraph could be read, "those same customers would prefer peanuts to anything else other than crackers"? I can see how it could be taken that way, and maybe that is how the survey was worded or something...
ReplyDeleteSimple. It's because statistics results can be bent in any way that the person presenting them wants.
ReplyDeleteKarnov,
ReplyDeleteAre you a statistician by training?
;-)
Ok, if the survey had two questions:
ReplyDelete1) Would you rather have a cracker or nothing at all?
2) Would you prefer a peanut to anything else (besides crackers)?
The order of preference could go
1)crackers, 2)peanuts, 3)everything else, 4)nothing at all, and the survey could still make sense. Even if this is what you were getting at, I still prefer Karnov's answer...
The junior employee argued that 1) Crackers are better than nothing and 2) Nothing is better than peanuts!
ReplyDelete