If we are to get anything useful from looking at the "persistence" data,
it may help to pick the data apart, so we are not comparing apples to
oranges.
There are four main hypotheses that need to be considered. Each of the
following "should" contribute to high persistence:
Before I looked at the data, I guessed that a school with a narrow choice
of majors would have serious persistence problems. A Caltech student who
gives up on quantum cosmology and decides instead to pursue 15th century
French literature is going to have to transfer out.
But the numbers don't support that guess. Apparently items (2) and (4)
outweigh items (1) and (3), at least in the examples I've looked at.
Some "boutique" schools with a limited choice of majors and rigorous
("drinking from the fire hose") classes have rather high persistence
numbers.
So far so good ... but there are still things about the data that I would
like to understand, but don't (yet) understand. For starters, how does
Haverford manage such a high persistence number? By any objective measure
I can find, Stanford is /more/ selective than Haverford, has impressive
amounts of money, and offers a flexible curriculum ... yet has a much lower
persistence number. I don't know how to explain this. Maybe there's a
simple explanation, or maybe not.
You could say that Haverford is anomalous, and that the basic trend
remains clearly visible: the more selective the school, the higher
its persistence number. Exceptions do not invalidate this simple
model. I would have to agree with that, except that I am not merely
trying to model the data; I am trying to /learn/ from the data, and
often the best learning comes from the outliers. A gold miner's
livelihood does not depend on the "average" amount of gold per unit
area on the earth; he digs where the gold is much, much more abundant
than average.
====================
We may learn even more by looking at the not-very-selective schools.
This is not some Lake Super-Wobegon situation where all the children
are in the top 5%. Most of the students are /not/ in the nations top
5%, and they have to go to college somewhere. For a not-very-selective
school to become highly selective doesn't solve the problem of how to
educate the ordinary student; it just shifts the problem somewhere else.
A highly selective school can do things that other schools cannot. So
there is a good chance that things we learn by studying the ultra-selective
schools will not be applicable elsewhere. So let's see what happens when
we look at more ordinary, mainstream schools.
Again the general trend is that more selective schools have a higher
persistence number. For example, you might naively think that UNC Chapel
Hill and UNC Charlotte would be comparable as to funding, rigor, etc.,
but Charlotte is less selective and has a lower persistence number ... by
a factor of three. Ouch.
I can find outliers, which beg for an explanation. For example, according
to the latest data, Sarah Lawrence College admitted almost everyone who
applied, and didn't even ask to see SAT or ACT scores ... yet their persistence
number was up there with UNC Chapel Hill. Wow. It's not like they gave up
on rigor. Nobody majors in Study Hall there.
I think this is worth pursuing, at least for a while. Maybe I'm chasing a
ghost, looking for something that's not worth finding, but maybe not.
I have a half-formulated hypothesis that persistence number have something
to do with student motivation. And that's relevant to this group, because
one of a teacher's main duties is to motivate the students. If the students
were all self motivated, they wouldn't need teachers; they would just go
read the textbooks for four years.
And motivation -- unlike selectivity -- is something where the whole community
could improve, if there are good ideas out there that some schools have
caught on to but others have not (yet).
What is "motivation", anyway? If you could sell it in a bottle you'd make a
fortune. I sure don't understand it. AFAICT it's not quite the same as
"school spirit", as exemplified by MIT seems to attract highly motivated
students, whereas their school motto is "IHTFP".
Michael E. identified a couple of special cases:
As for the student who stumbles into (or is pushed into) pre-med even though
he doesn't like math, physics, chemistry, or biology ... that's just bad
advising and bad counseling. The admissions committee should be able to
catch that during the application process. Secondarily, there should be
some outreach to the high-school guidance counselors, to tell them that
such a path is not going to end well. OTOH it may well be impossible to
talk sense to some such applicants.
I know less about the "sports crazy" applicants, but probably the same
words apply. SOMEBODY needs to tell them that there are very, very few
professional sports jobs out there. I forget the details, but I once
saw a study of this-or-that group of kids, 68% of whom planned on becoming
professional athletes. Only 1% of them actually succeeded, and that even
included the minor leagues and other not-very-rewarding jobs.
I've faced situations like this, where I had students who had zero chance
of long-term success. I advised them of this in no uncertain terms, and
even made them sign a statement for my files, because I didn't want to
be accused, later, of having held out false hopes or taken their money
under false pretenses. Most of them stayed with the program for a while
after that. It's amazing. I wish all students had that level of motivation
,,, or that such motivation could be refocused onto more reasonable
objectives.