recently i dropped an econometrics course because it was being taught with the
point of view of statistics as a black box. having become accustomed during 6
years of engineering studies to learning and understanding the analytical
concepts from the ground up (as it were), i found that i just wasn't satisfied to
do statistical economics any other way. when i contemplate a problem in
engineering or physics, my thought processes are informed by this knowledge of
every level of theory & practice, and as i look at the problem this way and that
in my mind i can check every level to satisfy myself that my analysis &
conclusions are supported by a firm theoretical foundation without any pockets of
uncertainty.
however, the approach i found in econometrics, of using statistical procedures
and distributions i barely understood and getting numbers out of a black box that
i was supposed to interpret, made me feel extremely uncomfortably analytically,
as though i was supposed to walk through quicksand where i couldn't see the
bottom instead of walking on a firm and familiar foundation as i have become
accustomed to do.
so i dropped the course because i just wasn't willing to proceed so blindly. i
concluded that i needed to study more statistics before i went any further.
however, it seems many folks who use statistics in the social sciences have math
phobia, and most curricula seem to not require the kind of educational foundation
that would enable students to validate black box output. to my mind, one needs
to do lots of problems which are analytically accurate and numerically simple,
and instructors should be prepared to provide very precise and detailed problem
solutions. in economics, i found a disturbingly lackadaisical attitude about the
need for the kind of assiduous attention to accurate procedures that i had come
to take for granted while studying engineering. setting up excel to perform the
common kinds of number-crunching is a way for the student to make sure s/he
understands what's going on, _if_ the instructor is prepared to provide the kind
of detailed feedback or problem solution that will enable the student to track
down every error.
i'm reminded of the TI30 calculator i used as an undergrad; about 3 times a year
it would give me an answer that was clearly bogus. because i had learned to
always be asking myself if the numbers made sense in context i was able to catch
these errors.
muriel
dennis roberts wrote:
> i would like to comment on this ... without getting anyone mad at me. i
> have heard this argument many times before but ... i think that if we
> promulgate this ... what it means is that we are not doing our students any
> favors ... i don't view some general stat package as a "specialist" package
> ... it is how people who do statistics work, work. it is comparable to
> saying that since everyone might have notepad on their machine, that that
> is the way they should do word processing. we need to alert students to
> general tools that are DESIGNED to do certain things ... and, if they
> become professionals in the field .. then they should know that sometimes
> you need to purchase "tools" for your work
>
> sure, excel can do (with its plugin modules) much of this stuff but, A) the
> algorithms for doing much of this are not very efficient nor, is their
> accuracy without conern, and B) excel is limited in many ways so, for many
> things one has to go to the REAL things anyway ... why don't start them on
> that path in the first place?
>
> software is cheap nowadays ... for example ... one can go to
> http://www.e-academy.com ... and download minitab) as one exmaple ... for
> $25 for 6 months ... or go to http://www.minitab.com ... and get it totally
> free for 30 days ... or buy good student editions of packagages for,
> reasonable prices OR, there are shareware packages that are pretty good ...
> and even online routines (like statlets) ... in this climate ... it seems
> even LESS of an argument that students might not have access to real
> packages .. they are all over the place and cheap.
>
> but, this is just my opinion ...
>
> At 06:46 PM 11/29/99 +0000, Graham D Smith wrote:
> >> I guess "nice" is in the eye of the beholder. I view this as
> >> another good reason NOT to use Excel for Statistics!
> >>
> >> Please use the right tool for the job.
> >>
> >> Jon Cryer
> >
> >Sometimes the right tool for the job is Excel. Most of my students will not
> >have access to specialist statistical software after they graduate. Although
> >Excel has many shortcomings, it is widely-available.
> >
> >
> >Dr Graham D. Smith
> >Psychology Division
> >School of Behavioural Studies
> >University College Northampton
> >Boughton Green Road
> >Northampton
> >NN2 7AL
> >
> >Tel (01604) 735500 Ext 2393
> >Email [EMAIL PROTECTED]
> >
> >
> >----- Original Message -----
> >From: Jon Cryer <[EMAIL PROTECTED]>
> >To: <[EMAIL PROTECTED]>
> >Sent: Monday, November 29, 1999 5:54 PM
> >Subject: Re: Rank Awful
> >
> >
> >> I guess "nice" is in the eye of the beholder. I view this as
> >> another good reason NOT to use Excel for Statistics!
> >>
> >> Please use the right tool for the job.
> >>
> >> Jon Cryer
> >>
> >> At 05:58 PM 11/26/99 +0100, you wrote:
> >> >Very nice solution.
> >> >It can be reduced even to the last part:
> >> >RANK(A1,A$1:A$6,1)+(COUNTIF(A$1:A$6,A1)-1)/2)
> >> >
> >> > Ivan
> >> >
> >> >> You can modify the rank() function using the following one:
> >> >>
> >> >> IF(COUNTIF(A$1:A$6,A1)=1, RANK(A1,A$1:A$6,1),
> >> >> RANK(A1,A$1:A$6,1)+(COUNTIF(A$1:A$6,A1)-1)/2)
> >> >>
> >> >> In this setting, the data range is A1:A6 (as the example you mentioned
> >in
> >> >> your email), and ranking is in the ascending order, e.g. smaller number
> >> >> gets smaller rank. You can type this function in cell B1, and copy the
> >> >> formula to B1:B6. Let me know if you need further info.
> >> >>
> >> >> Jay
> >> >>
> >> >
> >> >********************************************************
> >> > Ivan Zezula, Safarik University, Kosice
> >> >
> >> >
> >> _____________
> >> ------------------------------------------------- | \
> >> Jon Cryer [EMAIL PROTECTED] ( )
> >> Department of Statistics http://www.stat.uiowa.edu\ \_
> >University
> >> and Actuarial Science office 319-335-0819 \ * \ of
> >Iowa
> >> The University of Iowa dept. 319-335-0706 \ /
> >Hawkeyes
> >> Iowa City, IA 52242 FAX 319-335-3017 |____________ )
> >> ------------------------------------------------- V
> >>
> >
>
> ----------------------------------------------
> 208 Cedar Bldg., University Park, PA 16802
> AC 814-863-2401 Email mailto:[EMAIL PROTECTED]
> WWW: http://roberts.ed.psu.edu/users/droberts/drober~1.htm
> FAX: AC 814-863-1002
--
Any resemblance of any of the above opinions to anybody's official position is
completely coincidental.
Muriel Strand, P.E.
Air Resources Engineer
CA Air Resources Board
2020 L Street
Sacramento, CA 59814
916-324-9661
916-327-8524 (fax)
www.arb.ca.gov
