Latest Breaking News
In reply to the discussion: Human DNA Found In 66% Of Vegetarian Hot Dogs [View all]mahatmakanejeeves
(57,459 posts)Last edited Tue Oct 27, 2015, 11:27 AM - Edit history (3)
That information is missing, certainly from the USA Today McStory. Is it 1%? 2%? 3%?
Let's say the samplers went out and bought 3,000 packages of hot dogs. What we have been told is that 2% of those 3,000 packages contained human DNA. That's 60 packages. Of those 60 packages, 40 were vegetarian products.
What I do not know is if vegetarian products are overrepresented in the contaminated group, underrepresented in the contaminated group, or on a par in the contaminated group. If you went out and bought 3,000 packages of hot dogs at random, how many of those would be vegetarian products?
Are vegetarian hot dogs more likely to contain human DNA than regular hot dogs, less likely to contain human DNA than regular hot dogs, or as likely to contain human DNA than regular hot dogs?
Let's say vegetarian hot dogs are 5% of the market. I don't know; I'm just making up numbers. Of the 3,000 packages of hot dogs the samplers buy, 5%, or 150, are vegetarian, and 95%, or 2,850, are regular. The testing finds 2% of the total population, or 60, are contaminated. 40 of those are vegetarian, and 20 are regular. In terms of percent, 40 out of 150 packages of vegetarian hotdogs, or 27% of them, are contaminated. 20 out of 2,850 packages of regular hot dogs, or 0.7%, are contaminated. That would make vegetarian hot dogs 38 times as likely as regular hot dogs to be contaminated.
Did I do that right? Outraged statisticians, please feel free to point out my errors (beyond the obvious one, pretending to be a statistician).
I doubt that vegetarian hot dogs constitute as much as 5% of the market. Juggle the assumptions around, and the numbers change.
Also, are the results being skewed by false positives? I recall the paradox of false positives on drug tests. Give me a few minutes, and I'll find an account. Here we go:
False positive paradox