Alan,
I teach the introductory programming subject (Eiffel), first year.
The subject runs twice a year. In semester 1, I teach the
full-time class, straight out of high school. In the second semester, I
teach the part-time class many of who are professional programmers, who
have taught themselves or done a couple of non-university programming
subjects; that class also picks up the full-timers who failed first
semester.
Almost every semester I see smatterings of a bimodal distribution,
in each assignment (three assignments each semester), though it tends to
go away when all the marks are combined. I often see the same students
do the subject several times, so repeating the subject doesn't seem to
benefit them that much. In the full-time class, they all had excellent
scores in their end of high school exams, so they would all seem to be
reasonably intelligent ...
Over half my classes are Asians who see computing as a good
career where English is not essential (hah!), so there may well be a
language component to this pattern of data. Still, I see many native
English speakers who also never really seem to "get it".
I have a vague memory that IBM used to filter their cadets with
some form of test. Can any IBM alumni out there say more about this?
Robert Rist
Computing Sciences, University of Technology, Sydney
PO Box 123 Broadway, Sydney, NSW 2007 Australia
