At 4:04 pm +0000 30/11/98, Alan Blackwell wrote:
>> Executive summary: Students learning to program typically divide into two
>> populations: those who can easily learn, and those who can't. Teaching
>> techniques and curriculum innovation seem to have no effect. Are there any
>> effective ways of testing student aptitude before entering the course?
>
>Richard,
>
>Your observation was made specifically of a postgraduate conversion course.
>I suspect that this context might result in particularly exaggerated
>effects. In Cambridge we teach a postgraduate diploma conversion
>course in Computer Science, where the same thing is often noted. This
>seems to be because students join the course with widely differing
>motivations. Some students come with a maths or physics degree, in which
>they have done a substantial amount of programming, or even with a
>couple of years post-graduate experience in computing research, which
>they want to supplement with a formal qualification in computer science.
>Other students come from non-computing backgrounds, and use the diploma as
>an entree into the world of computing.
>
>It's clear that this will result in a bimodal distribution, and we do
>our best to structure the course in a way that caters to the varying
>requirements of these groups.
>
>Based on my experience of these students, and on your observations, I
>have the following points to make:
>
>1. Rather than students who "can learn" and "can't", all we are seeing
>is students who "have learnt" and "have yet to learn". If so, this is
>simply a matter of educational politics, rather than student aptitude.
>
>2. Is there any reason why these conversion courses don't attract a
>middle kind of student? If we assumed that programming knowledge or
>aptitude was normally distributed, this bimodal distribution might
>simply indicate that we are excluding the middle of the distribution
>when promoting or selecting for these courses.
>
>3. Does the same dynamic that I have observed also apply to undergraduates,
>or was Rob Rist also referring to students on a postgraduate conversion
>course?
The evidence presented was from a PG course. Exactly the same phenomenon
appears in UG courses. Our PG students are all non-programmers, bar one or
two; there is absolutely no correlation with prior 'mathematical' courses
such as engineering, chemistry or anti-correlation with non-maths ones like
politics or language. If it were just a matter of experience, if we were
seeing those with some background in the upper hump, and the others below,
I wouldn't have bothered the community with my question. Regretfully,
after twenty-five years of trying, I have to conclude that we must look
elsewhere for an explanation.
The middle students don't exist, at either PG or UG level. This isn't a
matter of those who have learnt and those who have yet to learn, I'm sorry
to have to tell you. The assumption that programming aptitude is normally
distributed would have to have some pretty tremendous evidence to overcome
the universal experience of programming teachers (and programmers'
employers) that it definitely isn't normally distributed.
Richard Bornat
