A couple of comments.
When a student leaves college and is not in a position (job) where he/she has
to use statistical formulas every day formulas are quickly forgotten. What
we (at least I do) expect is that they understand the basic principles of the
subject and can formulate ideas regarding the null and alternative hypothesis
(maybe not perfectly), the concepts of testing and such, the knowledge that
there are restrictions on the use of certain tests and routines and the
concepts regarding interpretation of the results of any test which is applied
to their data. Given that they will not be using the subject on a day-by-day
wouldn't it be better if they had created some form of reference file to help
them decide on what test to use and how to use it when faced with an
occasional problem? There are many statistical software programs around that
can easily be kept on a desktop PC or laptop to do the "donkey labor"
involved in the calculations. Many even help in the selection of the proper
test to use provided the user answers some simple questions. The software
becomes a tool. The selection of the proper test and the interpretation of
the results of the analysis become the job faced by the user. A carpenter
must know that a screw driver is used on screws and not a hammer. It would
be great if everyone could memorize everything we teach but stressing
memorization of formulas may not be worth the time it takes. I think other
parts of statistical analysis are more important.
How many of us have had to look back in out 'old' college texts to remember
how to do something? I know I have. Unfortunately in today's college
environment text books are VERY expensive (rather too expensive) and many
students, upon leaving the course, sell their texts. I'm sure that when I
attended college I felt the texts were expensive BUT I kept those I felt
might be important to me in the future along with the note book I created
while attending class. I'm not sure that students today keep their texts but
they can easily keep their notebook and 'crib' cards for future reference.
Another thing comes up when using statistical software taught at the
university. How to use it is fairly quickly forgotten unless the software is
used on a regular basis. It the use is complex requiring the user to go back
to the reference manuals each time it is needed (on an occasional basis) then
the software becomes less of a useful tool and more of a burden. The
software that is used in schools should be easily learned so that it can
quickly be put to use by those who have left school but need something on an
occasional basis.
We who teach tend to use, know, and depend on statistical software because we
use it everyday and because we often have to teach its use to students.
Picture however, a student who had to use SAS, SPSS, MiniTab or even Excel in
school who six or eight months after graduation picks up a magazine
containing information (research) in the field in which that student is
employed. He/she reads the article and decides to go over the analysis of
the data. Assume again that it involves a One-way ANOVA between groups.
Would you expect someone who used any of the above software in school to
remember the various commands needed by that software to make such a check on
the magazine data?
In the world outside of the college environment are individuals required to
use their memory of things that they had to memorize (or try to memorize)
while in school or are they expect to know how to look up information if
their memory is "fuzzy"?
Crib cards, if the instructor helps in the logic of their creation, might be
the most valuable took that the student carries from the class. The use of a
software package that doesn't require programming or commands or special
entry methods might be the second most important.
Dr. Robert C. Knodt
4949 Samish Way, #31
Bellingham, WA 98226
[EMAIL PROTECTED]
The Law of Gravity says, "No fair jumping up without coming down."
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
