Lots of press interest in this paper here in the UK. Interesting paper
though......
The good thing about it is that it accepts that it has found an
"association" and doesn't try to imply (like many other papers) that there
is anything stronger. It also accepts that more research is needed....
The problem with "association" papers is that the press gets hold of them
and writes things like "Link found between MMR and autism" (which people
read as "MMR causes autism") and then the whole thing kicks off.
There are many criteria which need to be fulfilled before a link and
causation are proven.
Lets not get started on Dr "MMR" Wakefield's stats and logic anyway- his
double hit phenomenon was based upon retrospective observations on one
patient!
The sleep/mortality study does have a very large sample size which is always
good. However, I always worry about studies which look at so many variables
from the very beginning. It's not so much a hypothesis proving exercise,
more of a fishing trip...
Thinking of the association problem, I remember my teacher in critical
reading telling us of the association between increased stork numbers and
increased new born babies after WWII. It even proves a long held belief :)

Ben

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads