On Apr 18, 2012, at 12:31 PM, Jeff Newmiller wrote:
I think that mostly avoiding the statistics and matrix capabilities
is wise. You might want to (re-)read Burns' article on Spreadsheet
Addiction for help in justifying the effort required to learn R.
In that vein, there is a classic experiment where a small ball is
rolled down an inclined pane and the time required to roll various
distances is measured.
I'll say it's classic. First done by Galileo:
http://www.u-picardie.fr/~dellis/Documents/PhysicsEducation/Reconstruction%20of%20Galileo%20Galilei.pdf
One way to investigate fitting this data is to square the time in
the spreadsheet.
Another way is to examine first and second differences in distances
reached after successive equal intervals. The diff() function is
rather handy in this effort.
Back to the matter at hand, ... I see no convincing reason to consider
R any more complex as a computer language than is Logo. The suggestion
to use color in graphics output seems to be in accord with what I have
seen as far as pedagogic recommendations for using Logo as a teaching
platform.
--
David Winsemius.
(The other is to use the built-in polynomial regression.) If there
is a missing value in the input time, the squared cell will be zero.
You can overcome this by manually putting an =NA() in the missing
cell, but that is tedious when there is lots of data, and it gets
even more tedious when you want to throw out the whole data record
while remembering which records were used in the final analysis. R
allows this and similar steps to be automated.
I also think running some examples of plots like tiled layouts,
colored maps, boxplots, or 3d interactive cloud graphs may provide
good brainstorming material for data representation.
---------------------------------------------------------------------------
Jeff Newmiller The ..... ..... Go
Live...
DCN:<jdnew...@dcn.davis.ca.us> Basics: ##.#. ##.#.
Live Go...
Live: OO#.. Dead: OO#..
Playing
Research Engineer (Solar/Batteries O.O#. #.O#. with
/Software/Embedded Controllers) .OO#. .OO#.
rocks...1k
---------------------------------------------------------------------------
Sent from my phone. Please excuse my brevity.
"Christopher W. Ryan" <cr...@binghamton.edu> wrote:
Thanks all for the excellent thought-provoking comments.
I want to clarify that these students are, for good or for ill,
already
doing all these analytical and graphical things for their projects.
They
are doing them with Excel and SPSS. One of my goals would be to teach
them how they can be done (and I think done better) in R. Better for
many reasons, not least of which is the reproducibility offered by
lines
of saved code.
It seems that many (not all) on the list agree with the science
teachers
that R is too difficult for high school students. Is R intrinsically
more difficult to learn than commercial spreadsheet software? If so,
why? Or is the issue that it is difficult to change to R after many
years experience in the mind-set of spreadsheets? If a child was
"brought up" on R for math/stats, in a developmentally progressive
way,
instead of Excel or a graphing calculator, would he/she perceive it
as
difficult?
Are the intrinsic cognitive differences between high schoolers,
college
students, and graduate students substantial enough to explain why the
last can learn R and the first can't? Or is it a matter of exposure,
opportunity, etc?
Indrajit, I'm curious: given your preference for hand-drawn graphs
for
learners (a very good point), why is Excel "fine" but R not?
At any rate, I should probably migrate this thread over to the
Teaching
SIG listserve, which I didn't know about before.
Thanks again.
--Chris
Christopher W. Ryan, MD
SUNY Upstate Medical University Clinical Campus at Binghamton
425 Robinson Street, Binghamton, NY 13904
cryanatbinghamtondotedu
"Observation is a more powerful force than you could possibly reckon.
The invisible, the overlooked, and the unobserved are the most in
danger
of reaching the end of the spectrum. They lose the last of their
light.
From there, anything can happen . . ." [God, in "Joan of Arcadia,"
episode entitled, "The Uncertainty Principle."]
Bert Gunter wrote:
<...snipped>
I anticipate keeping things very simple:
--objects and the fact that there is stuff inside them. str(),
head(), tail()
--how to get data into R
--dataframes, as I imagine they will mostly be using single,
"rectangular" datasets
--a lot of graphics (I can't imagine that plot(force,
acceleration)
is beyond a high-schooler's capability.)
--simple descriptive statistics
--maybe t-tests, chi-square tests, and simple linear regression.
I have some experience in this and would have to agree with
Indrajit
that this is not a good idea.
When I tried to teach R to a high school student it was not very
successful. Certainly based on that experience the list above is
way
too complex. Don't teach anything on that list at all. The number
of
concepts involved in that is simply overwhelming.
Oh amen amen!
I'd go farther: It's overwhelming for college students.
Farther yet: I've met very few scientists and engineers who
understand
what a standard deviation is. Fewer still who understand the
difference between a sample statistic and a population parameter for
which it's an estimate.
This approach to "basic" statistics is (imho) symptomatic of why our
discipline is so widely disliked and misunderstood.
Cheers,
Bert
Also avoid teaching
anything that requires complex installation if you want them to be
able to carry it forward by themselves.
I would expect the reaction would be that most will have no
interest
and the ones that do will be frustrated by the large number of
concepts needed to get going.
The only part that seemed to trigger any interest was when I showed
the large list of colors available in colors() and then playing
with
inserting different colors in:
colors()
plot(1:5, col = "violetred")
Assuming you are committed to this and go ahead, I would divide it
into two parts:
1. a graphics demo -- make it clear its a demonstration so they
have
an appreciation of what is possible and you are not actually
teaching
anything in this portion.
2. Teach them how to install R, run the above two commands
(substituting in different colors), how to exit and point out that
there are many tutorials in:
http://cran.r-project.org/other-docs.html
and they can pick one they like (since the official documents will
be
over their head).
If you do that then perhaps a small number will have sufficient
interest to try it some more at home but I wouldn't be surprised if
none do and that most or all would prefer something with more
immediate gratification.
--
Statistics & Software Consulting
GKX Group, GKX Associates Inc.
tel: 1-877-GKX-GROUP
email: ggrothendieck at gmail.com
David Winsemius, MD
West Hartford, CT
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
