Consider that if I go out and ask three criminals in a penitentiary,  do you like big breasts or do you prefer small breasts and two men man replied to me, "I don't like breasts at all,  I like dicks", while the third man tells me he prefers small breasts. Now I could go out and tell the entire world that two out of three men prefer dicks to women who have small breasts.   Obviously this is not the case.  My poll suffers from at least two defects.   First, the sample size is too small.  Asking just three men the answer to just about anything is to rely on a sample size of just three which is so small as to be statistically irrelevant.  The second huge flaw in my poll or sample group is that I have preselected a prison, and let's face it, most people in a prison are not normal.  So even if I were to go out to 100 prisons and ask 20 men in each prison the same question the results in my poll are still suspect.  I can make the case that the results of my poll are accurate only to the extent that they reflects what criminals living in prisons think about breasts and dicks and that they by no means reflect the tastes of non criminals who do not live in prisons.

Now......up to now over 2400 people have read my first article, "Tits" and had a chance to see that there is a poll that they are invited to participate in.  Obviously that number 2400 is statistically significant.  The sample size is certainly large enough, but what is it exactly that we are sampling here?  Thankfully I have a hit counter on the web page entitled "Tits" that shows where all the visitors to this page come from.