In My Opinion, Average BPEL is:

If a sample of 100 is equally valid whether the population is 100,000 or 100 trillion, then why are so many studies conducted with thousands of samples? Even “random” political polls conducted with between 500-1,500 respondents were proved by this last election to have been so inept. Why? Because the sampling was too low.

I would offer a simple example. When the FDA finally releases a drug for public use, it is only after thousands of subjects have been tested. Yet, even after so large a sampling, they’ve been known to pull a drug off the market years later when, after hundreds of thousands of patients have used it, harmful side-effects begin to develop (even with new users). If your theory is inviolable (as pertaining to human physiology), then why weren’t these issues known after only 100 test subjects?

I’m afraid you’re confusing the discipline of statistics (as applied to mathematical probabilities, physics, etc.) with the more “unpredictable” field of human physiology.

Consider that the largest penis ever measured was 33.5 cm (about 13.2”). If you consider the low end to be “retractable penis,” you’d have to allow maybe a 0.5” EL (or less). That gives a range of 12.7 inches. If you chose to measure in even 1/8” increments, you’re still talking about a possible 100 different measurements (providing that no subject is found longer than 33.5 cm).

And that doesn’t even address the frequency of recurrence. If you wanted to assemble a thorough graph in which you charted at least 10 instances of each of those 100 measurements (with the exception, of course, of the “longest”), you’d have to allow that in order to get 10 for some sizes, you’d need 20, 30, 50, 100 or more measurements of more “common” sizes. Even a low estimate of such a compilation would involve thousands of subjects.

Furthermore, if you conducted such a survey in a quite homogenous society (like Israel, Japan, etc.) vs. a pluralistic society (like the U.S.), you’re furthering skewing the results due to ethnic/racial peculiarities.

It’s not my lack of understanding of the Central Limit Theory that’s at issue, but rather your overzealous application to it regarding this issue - a mistake, by the way, that the professionals researchers of the established institutions do not make.

You ask lots of good questions here, Wadzilla, and make some good points that I would not contest (nor have I contested). Unfortunately I do not have time to provide all the explanations you seek here at Thunder’s Place; typing takes a long time, as you well know. If you ever happen to be visiting the "established institution" at which I am a professional researcher, however, feel free to attend a lecture ;) . Rest assured I am not "confused" about this stuff (nor do you seem to be confused about most of it). This appears to be a good book on the topic, if you’re curious, and it can probably be found a lot cheaper on Half.com or the like. You’re obviously an extremely bright guy and you’d probably learn a lot from it, even if it’s aimed mainly at medical folks.

One other thing: I am not saying that medical scientists have no clue what they’re doing. You are the one who criticized the penis-size medical studies in the original thread because you thought their sample sizes were too small. Sometimes studies do employ too few subjects, yielding null results, statistically, when there really was an effect to be found, and likewise, sometimes studies employ way more subjects than necessary, wasting tax dollars so that the headlines look impressive. But typically things are done right. I’m just defending the status quo, like the good reactionary I am :) .