Total Pageviews

Sunday, July 19, 2020

How to study a study

I'd like to read a scientific
study or two every week, 
but many are not free.

I retired in 
January 2005,
at age 51, giving up 
a stressful white
collar job,
to become 
a professional
lazy bum,
... and mandatory 
cheapskate.

So I'd be upset
if I had purchased
any scientific study
that provided nothing 
of value.

First of all, roughly
half of study results 
can't be replicated !

A 'slight' problem !

And most of the
climate "studies"
are nothing more 
than wild guesses
of the future climate.

Always bad news
coming, it seems !

You'd have to wait 
many decades, 
or even a century, 
to find out
if the predictions 
were right.

They are rarely right.

We've been hearing these
climate related predictions
since the 1970's -- long 
enough to realize their 
batting average is near
zero.

I don't see wild guess,
long term predictions
as real science studies.


Ignoring that 
HUGE problem
for now, three types 
of studies that 
annoy me the most
( three examples follow
in the next three posts,
(A), (B), and (C) ):


(A) 
A study that claims
the predictions 
are right, even if 
they don't match 
the subsequent 
measurements.


(B) 
A study that states 
the obvious.


(C)
A study that 
agrees with 
so many similar 
past studies,
that it adds little 
to anyone's
knowledge.