Totally agree with this. To my knowledge, there's only been one good study in terms of the sample set -- the Italian conscript study (Ponchietti et al.) In parts of the world still backwards enough to have conscription (I know.. save it for the politics page
) everyone at 18 will go in and have their military physical. This is a little past the age when development will have finished for nearly everyone, so it's probably the best age to take the study at since you'll be measuring young, healthy dicks and not ones damaged by decades of smoking and obesity. They're also way more fit than Americans, and again being young helps with that, so this really lessens the impact of bone-pressed vs non-bone-pressed since there's not much of a fat pad there to begin with. They measured complete stretched length (not a 'gentle' stretch, but stretching to the maximal distance after 3 warm-up stretches), which the Wessels study proved is an extremely accurate predictor of erect length (generally only around 0.4cm shorter stretched.) This means there's no question of whether the guys were just sporting partial chubbies or had a curvature that wasn't measured or anything like that.
They did 3300 people (for statistical measurements of a population, that is a hell of a lot.) The mean stretched length? 12.5cm! That's just over 4.9" long. If you add some length back on there to predict an erect length based on stretched numbers, that's still only around 5.1", which is what the Wessels study got years ago and most medically supervised, measured studies that didn't just go around asking people if they wanted their dick measured got.
There are two big factors that often bump the mean number up: self-selection, and self-measurement. Grouping them by these factors I've noticed it's about:
no self-selection or measurement: 5.1"
self-selection only: ~5.6"
self-measurement only: ~6"
self-selection and self-measurement: ~6.5"
anonymous internet survey: the size of a 2L bottle of coke.
The standard deviation on the Italian study is only 2.5cm, so the difference between the general population, and wishful-thinking exhibitionists, is about 1.6 SDs. I know it's not quite a normal distribution, but that puts the high "average" at about the top ~8% of the population. This also means there's enough overlap that between two guys with exactly the same 5.8" dick, one might believe that he's a significantly bigger than most guys, while the other would think he's smaller.