posted
This is true; the 99% confidence interval for the mean around the mean of a sample that size is breathtakingly small. In fact, its relatively easy for us to approximatet an upper bound:
Using three as s, and looking for the 99% confidence interval for a sample of size 15,000 (upper critical value 2.576), we find the confidence interval would be about .06 (years).
Unless the sample for the overall population is significantly smaller (which would be quite odd), that the difference is significant is not in doubt. Including Mormon non-smokers from Utah will have negligible effect since we know their population mean is similar to the national population mean, and including non-Mormon non-smokers from Utah only serves to narrow the gap, since we know their population mean is less than the national population mean.
As for what the statistics actually mean, that's certainly up in the air. But there is a significant difference between the populations, no doubt at all.
Posts: 15770 | Registered: Dec 2001
| IP: Logged |
posted
Wow enochville. I bet you never knew that now that you left the Mormon church you just whacked 5 years off your lifespan!
Posts: 1412 | Registered: Oct 2005
| IP: Logged |