[R] follow up on teaching R to high school students

[Christopher W. Ryan]

Some of you may recall that a few months ago I solicited advice and
opinions on both R-help and R-sig-teaching about an "introduction to R"
workshop I intended to present to a class of high school students
enrolled in a 3-year longitudinal science research class.  Much lively
discussion ensued. Well, I did the workshop on 8 November, and thought
I'd give some follow-up as to how it went, in case anyone was interested.

The class consisted of about 20 kids, grades 10-12. Each was at a
computer with R installed. It wasn't clear that the students would be
able to install packages, given the school's network setup, so I
confined my plans to base R. We ran for 5 hours, with about a 40-minute
break for lunch. Their science teacher was present for the entire time,
and the school's IT person attended about half of it.  They were both
very helpful, both in preparation and in execution.

I conducted an (utterly arbitrary and unvalidated) online survey among
the students a couple weeks in advance, to gauge their familiarity with
what I called "technical computing," i.e. anything beyond commonplace
word-processing, spreadsheets, web surfing, and social media. The
questions were:

1. What operating systems do you know how to work in? Check all that apply.
   Windows 19
   Mac OS X 12
   Linux 2
   others 0

2. Do you have a favorite text editor?
   Yes 5
   No 7
   I don't know what a text editor is 7

3. Do you use a two-pane file manager?
   Yes 1
   No 6
   I don't know what a two-pane file manager is 12

4. Have you written programs in any computer language?
   Yes 4
   No 11
   I don't know 4
   (the specific languages cited included Basic, Java, Javascript, Ruby,
 C++, Python, MS-DOS command prompt batch files.)

I had my "lesson plan" all laid out in an org-mode file, from which I
typed code into an R console projected on the screen. The students
followed my steps initially, and then broadened out to some
experimentation as the day went on. A couple students were quite skilled
at working ahead, while others struggled a bit, but everyone was
eventually able to get the desired results.  They were generally very
engaged, interactive, and enthusiastic. No one left, except for the odd
music lesson here and there.  Overall, we had a lot of fun.

I tried to go pretty slowly. I prepared much more that we had time to
cover. I emphasized graphics. I did not get into inferential statistics
or hypothesis testing at all, despite their eagerness to "do a t-test"
and such. Maybe that will come at a future session, if we do one.

In general, topics we covered included:

vectorized mathematics (what I called "bulk math")
generating sequences
(meant to do logical conditions here, but skipped it inadvertantly)
drawing random samples
different kinds of objects (we limited ourselves to scalars, vectors,
dataframes; character, numeric, and factor)
levels of measurement: nominal, ordinal, interval/ratio
exploring objects: str(), head(), tail(), class(), summary()
using in-built data sets provided with R
general principles of good data entry and storage, and the virtues of
plain text. Went over read.table (I meant to do more with reading data
into R, but ran out of time. I sent simple instructions for the foreign
package and read.spss() to their teacher after the fact, since up until
now they had been using SPSS a lot, and several of their data sets were
in that format.)
graphs: boxplots, scatterplots, stripcharts, scatterplot matrices, and
coplots (they liked that last one a lot).  Also some graphical
parameters: type=, main=, sub=, col=, xlim=, ylim=, and pch=

Comments to teacher over the subsequent couple of days included:
"This should be taught in high school." "I got to see data for the first
time in a different way." "I had the most fun when I realized I could
play around with the program." (Of course, any less-than-positive
comments, the students (or their teacher) may have kept to themselves
out of courtesy to me!)

A post-workshop survey of the students has just begun, basically asking
them if they have downloaded R on their own computers, and whether they
have used it in any way since the workshop.

--Chris
-- 
Christopher W. Ryan, MD, MS
SUNY Upstate Medical University Clinical Campus at Binghamton
425 Robinson Street, Binghamton, NY  13904
cryanatbinghamtondotedu