The weirdness of statistics

[torquill]

If the predicted frequency of gluten-intolerant individuals (given current data) is 1 in every 250 people... How is it that I know so many of them?

I swear, I know at least eight diagnosed celiacs off the top of my head, which means I missed a few (there's a massive number of undiagnosed ones out there). There's at least one and often more of the confirmed celiacs at most events I attend, and that's not even counting the wheat-sensitive people like me. What gives?

Comments

[luna-torquill.livejournal.com]

There's a flaw in that logic -- you're assuming that the diagnosis rate, even out here, is 100%. I'd guess it's probably more like 50% at best, possibly lower. For many people, there are no symptoms other than vague ones, maybe not even obviously digestive in nature. It takes a blood test or biopsy to know.

And while I know a decent number of people, I am not regularly in contact with even 2000, let alone 4000 (assuming the 50% diagnosis rate). I'd guess I regularly run into around 200 people tops, and know another 250 through "oh, I've heard of them". My friends circle has shrunk dramatically from what it once was, and I don't socialize often.

So by that, I should (statistically) know two celiacs, who may or may not have been diagnosed. I know eight diagnosed and at least another four to six who haven't been, by extension. That's what's making me go "wtf?"