Let's talk about the data itself a bit first:
Microsized sizes are not included in these studies, so there's no way they could be "skewing" the average to a size lower than it really is. The probability there's more people on the bigger side of the spectrum is also not probable because there's literally too little room between 5" and 0" to lower the average as much as to counteract sizes ranging from 7" to 14" unless there were the same or more people at this side of the spectrum. As data tells us, it's likely for it to be mostly evenly distributed.
Alright, with that out of the way, you might say that when you saw these sizes wherever that they looked bigger than they really were. Well, I provide you the reason: perspective. Have a look at this (safe for work) image gallery with three paper rolls in sizes 7"x5", 7.5"x5.5" and 8"x6" (already huge!) compared to nearby objects using only foreshortening. See?
You normally view yourself from a top-down point of view, others view it in front of you, angled a bit to either side or up/down. That is enough of a perspective change to visually alter the size your member. You can only compare it to elongated parts of your body due to the distance, they have a direct view of the situation. Lastly, if you still don't believe it, do the same as the image gallery above, create a paper roll or anything cylindrical with your measurements and try to compare it a bit better, or even use similar objects as reference.
However, that is not the only trick the "advertising" industry knows, know how in most normal magazines you generally see unrealistic body ideals? A lot of it comes from image editing, of course, but the rest come from lightning effects, specific angles, shorter/taller people and other various tricks. If you grab a small person with a regular sized member, but don't put enough information to tell a person's actual size, it might look like they're an average-sized person with a huge member instead.
This video from Vsauce on YouTube talks about a similar yet somewhat unrelated effect, the "Moon Terminator Illusion" where people perceive distance wrongly without the right context in front of them. Skip to 2:17 if you don't have much time (and watch until 4:19).
Alright, about the measurements themselves, trust me, anything 6"x5" (15cm by 13cm) and beyond is already well above average. Most claims at or above 10" (25cm) are false, it'd be incredibly rare for you to find one even online. Not impossible as there certainly will be people of that size, but not probable. Most measure incorrectly (accidentally or not) or round up their numbers a bit too much. And because of the comparison issue mentioned earlier, people are also bad at guessing which size your member really is, often overestimating it (the so called "girl inches" vs. "real inches", but it can happen to anyone).