[plt-scheme] HTDP - evidently not for everyone.

This paper makes the rounds every few months.

It is certainly entertaining.

The "test" they propose strikes me as being as much about particular
vagaries as about general principles.

The statistics here is weak, and there is little one can meaningfully
conclude from their "longitudinal" component.  (I use scare-quotes
because of how little longitude there is.)

Their literature review is conveniently weak; their comments about
Wason and Johnson-Laird, for instance, miss about thirty years of
follow-on research, some of which offers perfectly good explanations
for those original findings.  (This is literature we happen to follow
in our security research, since similar questions arise when examining
whether and how people can formally construct security policies --
which are also program-like statements.)

Finally, it is unclear what the point is.  Should we administer the
exam and toss out all the students who fail it?  That would certainly
be a good idea.  For that matter, we can clearly administer this exam
earlier (say at age 14 or 15, when students are old enough to clearly
understand the problem statement).  Should we suggest that students
who fail this (or a similar test) skip algebra entirely?  That way we
can save them from wasting their time in a technical discipline.  But
maybe one could find similar "predictive" tests for writing; maybe
those students should be excused from writing courses.  Even better:
it ought to be possible to find such a test for reading (after all,
natural languages are also meaningless formal systems with arbitrary
and inconsistent rules).  We could keep such students out of school
entirely, to save everyone's time!

Personally, I do put a lot of stock into some of the material on
cognitive dimensions and mental models.  I am quite willing to believe
that our brains are wired differently.  But I view these as
challenges, not as insurmountable obstacles.

Shriram