News for hackers. Stuff that matters.
Re:Why is the news media comfortable with lying about science? by roundsquare on Friday January 15, 2010 @04:49AM (#1054588)
It strikes me that for a condition that is diagnosed in 1 out of 120 children, a population of 300 children is a bit sparse. Is that blasphemy?

In fact, you don't need large samples to get significant results. If there is a strong effect, even a very small sample will show it. If the effect is small, you will need a larger sample.

300 is, in fact, a pretty large sample. I haven't read the full study, but if its a simple random sample, the an effect would have to be quite weak indeed to not show up (in all probability).

There are no comments on this comment yet.