From: Kuiper, James D. (firstname.lastname@example.org)
Date: Fri Jul 26 1996 - 12:56:36 EDT
> One thing
> that impressed me as I learned more about statistics was the degree to which
> statistics are misused by non-mathematicians. I would advise linguists to
> exercise caution when tempted to draw conclusions based on statistical
I concur. I have math & statistics degrees and work in the statistics
department of a major medical institution. One of the hot issues within
our profession these days is the ever increasing use (misuse) of statistics
by people who don't know enough about what they are doing. I little knowledge
is very dangerous. The American Statistical Association, which meets next
week, is considering some sort of certification procedure to protect the
integrity of the profession.
A B.S., M.S., or PhD traditionally was enough to demonstrate competence, but
the discussion of certification demonstrates the growing worry and lack of
control felt within our profession.
We are all familiar with the explosion of junk science and dubious studies
within the popular media these days.
Anyway, regarding greek studies keep these tips in mind.
1. If you need some statistical analysis done consult a competent statistician.
I have a B.S. and they don't let me fly solo around here unless things
are very busy.
2. Be aware that there are many ways to misunderstand and misuse statistics.
Not everything is as it appears. For example, the number of clergy in
America is positively correlated with the amount of alcohol consumption.
These coincident occurences may or may not be related by cause and effect.
3. Statistics is often an attempt to quantify something that is very difficult
to quantify. Eg. someone's level of fear or anxiety.
Consequently, how one does the quantifying makes all the difference in
4. The meaning of a particular result may not be what you think it is.
The fact that I find an outlier in a particular sample may or may not
be a big deal. It depends on what is being sampled.
5. You may find something "significant" in your analysis. Significance is
a technical statistical term. This may not, however, tell you what you
think it does. If Paul uses "significantly" different vocabulary in
one letter compared to another it isn't the last word on authorship. The
real differences may be due to his purpose in writing and the subjects
he deals with. Again, interpretation of results is something you must
be careful with.
6. A great little book is titled "How To Lie With Statistics" written by
Darrell Huff. It covers many of these issues at an understandable level.
I have to go but if I think of more later I'll put some more together.
If you have more specific issues to discuss it would help us narrow things
down. The person with the Corinthian question did just that.
Yours with 95% confidence,
This archive was generated by hypermail 2.1.4 : Sat Apr 20 2002 - 15:37:46 EDT