("Matt Barrow" wrote)
Why is it bogus? The EPA claimed that 3,000 Americans die annually from
secondhand smoke, but there was a problem. They couldn't come up with that
conclusion using the standard statistical 95 percent confidence interval.
They lowered their study's confidence interval to 90 percent. That has the
effect of doubling the margin of error and doubling the probability that
mere chance explains those 3,000 deaths.


Let's not quibble, you went from 5% to 10%. Ok, I'll call that doubling the
margin of error.

They're at 90 (friggin') % for crying out loud - down only (10%) from 95%
.....confidence.

Also...
"That has the effect of doubling the margin of error and doubling the
probability that mere chance explains those 3,000 deaths."

Word games! Double-Plus Ungood. Yuck!

Again, ...they're at 90 (friggin') % for crying out loud.

Let's see if I can do those numbers in my head, hmm...that's 9 out of 10
case.

Sure, "mere chance" plays a role in some of the deaths, but a 9 out of 10
CONFIDENCE LEVEL IN THE REST. HELLO?

BTW, 90% is the statistical minimum. So we might be hanging around the 93.8%
mark. I'm just saying...

I might agree with your bigger point, but this argument, on its own, is
disingenuous - a sales job.


Montblack