I’m going to go a tad more scientific, but I think Jawbone here has the correct approach. Even though I think he is understating the relevance of making the deviation needed somewhat unreasonably large. I am pretty sure I understand why he is doing it.
In order to consider big vs small, you have to first get through Average.
If Average is both a number and a range, then I would suggest that there are two ways to consider it. One is that anything bigger is big, and anything smaller is small. That however, ignores the distribution, so you could use the SD to extend the range of average one SD up and down. This has the benefit that it includes the majority of penises. So some people are unwilling to use the approximate consensus around NBPEL being around 5.1 inches or so. It varies but that seems to be roughly the number. And the SD I believe was something around .8 or .9 but probably less than 1 inch. So I would say from this first approach, that anything over 6.1 or 6.2 NBPEL is big, while anything under 4.2 or 4.3 is small.
Despite my faith in the stats of the Cancun study (assuming the data is BPEL and not NBPEL which to me is uncomfortable because then it fails to jibe with practically any of the real life surveys), I’d have to say the SD is rock solid. And I think that they said 5.9 with a .8 SD. So by that metric, then anything under 5.1BPEL is small while anything over 6.7 is big.
However, another way to look at is with the 2 sigma approach, in which the top and bottom 5% or so are the criteria for inclusion. If you are 2 SD away from the median, then you are big or small. So here is where Jawbone has made a pretty good estimate. If 5.1 NBPEL were the median, then to get to 2 sigma, you need 6.7NBPEL. I think from what I have seen that this is pretty big. However, it doesn’t seem necessarily as big as 3.5 NBPEL does to be small. However, it puts a more concrete threshhold for big vs small in place, because it is less likely to be the subject of debate. If you’re in the top 5% it’s not really arguable that you’re not big.
So, taking into account the idea that there is a discrepancy in the agreement/consensus regarding Cancun vs UCSF, I think it’s safe to basically average them together. They both yield fairly close answers. I know for a fact that the 5.1 figure from UCSF is NBPEL so I’m going to keep my answer on the median there. However, I’m going to use the SD from Cancun because it’s a larger sample. So, My answer is 5.1 NBP +/- 1.6 yields big vs small and in such a large margin as to be indisputable.
4/2008 Bpel 6.50, Beg 5.5, Mseg 4.9
6/2008 Bpel 6.75, Beg 5.5, Mseg 5.1
9/2008 Bpel 7.00, Beg 5.5, Mseg 5.1
Last edited by Fantom : 06-21-2010 at .