On Fri, 22 Jan 2016 18:02:50 -0700, Ashton Crusher wrote:

It's a common statistical error (often done on purpose I think)
to get a politically desirable result.

Very often done, on purpose!

That's why the scientific method requires proof and logic and
the ability to reproduce (remember cold fusion anyone)?

Those studies that purport to show cellphone use as distracting
as, say, drunk driving hold sway.

They're useful data (if they're done well), but, their arguments
can't possibly be supported by the facts.

We all know, by now, that there are no accidents caused by cellphone
use over and above the accidents that are caused by all the other
factors involved in driving.

So, if cellphone use truly was as dangerous as those drunk-driving
analogies are, then there would be an increase in the accident rate
so huge that it would be impossible to hide by other factors (alien
manipulation included).

The fact that the accidents aren't occurring shows how those who
do those studies are looking in vitro, and non in situ.

In vitro studies are great; but they don't necessarily show what
happens in situ.

This study *did* look at in situ results, and what it found was
odd. Very odd.

First order = no greater number of accidents
Second order = no greater number of hospitalizations (injuries)
Third order = greater length of hospitalization (more severe injuries)

I agree. It's not intuitive.
Most people jump to conclusions (i.e., most people are idiots).

But it's hard to jump to a conclusion on this one.
Even if you are intelligent.