🚀 go-pugleaf

In article <1994Oct22.224511.696@rpp386>,
John F. Haugh II <jfh@rpp386.cactus.org> wrote:
>In article <Cy11nu.IJ2@txnews.amd.com> ricks@dvorak.amd.com (Rick Shank) writes:
>>Excuse me John but do you mind if I fix this?
>>
>>The
>>>science of statistics tells us that if you select a reasonably sized
>>>sample at random, that the properties of that sample PROBABLY apply to the
>>>entire population as a whole.
>
>Yes, but the probability that the sample is representative of the entire
>population is high.  Or put another way, it is unlikely that the
>conclusions reached as a result of that survey are wrong.
>-- 



I believe that is exactly what I meant.  The way it was originally written,
it was stated as an absolute rather than a likelihood.  The other problem
with statistics is that they are almost always used to state as fact, things
which are not valid due to the test/sample/data.  For example:  If a survey
showed most crimes were committed with a handgun, removal of handguns would
eliminate crime.  What would really happen is the criminal would find another
weapon.  Also, statistics only show trends.  They are not absolutes.