25 Feb 2008 ............... Length about 1,000 words (7,000 bytes).
(Document started on 18 Oct 2007.)
This is a WWW document maintained by
Steve Draper, installed at http://www.psy.gla.ac.uk/~steve/localed/tintopolicy.html.
You may copy it.
How to refer to it.
Web site logical path:
Best approach to dropout at Glasgow University
Department of Psychology,
University of Glasgow.
In my provisional opinion, at GU (not for all HEIs), we should adopt these
strands of practical policy.
- Calculate a sensible, not baseless, benchmark.
Estimate what the right comparison or benchmark for GU for dropout should
be. The Smith & Naylor (2001) paper establishes convincingly that the
government measures are not suitable benchmarks either as published in
league tables or in comparing GU to other Russell group HEIs.
- Use the Smith & Naylor (2001) corrected rankings (if we can
get the names of the HEIs), select out the Russell group, look at those.
- Other factors like 3 year vs. 4 year; average attainment of accepted
students incoming; average social class etc. are all reasons that crude direct
comparisons are not appropriate. In particular, GU does much, much better
than many other Russell group HEIs at widening participation, and this is well
known to be in tension with dropout rates. At the very least, should look at
the two together in the same table.
- Ignore comparative benchmarks, just attempt improvement.
(I.e. use benchmark measurements only as comparisons between past and future
performances of our HEI, and not between our performance and other HEIs.)
From an applied educational viewpoint, we could probably do better than we do
now, and should explore this and measure whether we improve.
- Do not treat retention as an independent objective, but as twinned with
widening participation. If we really wanted to improve retention in itself,
our best move would be to exclude all "widening participation" students and
reduce our performance on this to the poor levels most other Russell group
HEIs show: our retention would improve immediately. If we don't want to do
that, then any policy objective should say so. This might entail:
- State the policy as achieving both widened participation AND higher
retention, or at least a higher combination of the two.
- All tables showing performance should always show measures of both these
together. This will remind us of the tradeoff, and remind our competitors of
their poor WP performance.
- Develop modified measures that illuminate this e.g. show retention
figures both for all students and for the subset of our students that don't
count in WP figures.
- Ignore retention as a policy initiative in itself, focus on teaching
My general view, as the short
list of proven successes
implies, and based on theory and various other things, is that the best
general approach at GU for reducing dropout without abandoning widening
participation is to improve teaching quality.
I.e. a policy of driving up teaching quality would also drive up retention
rates, particularly if directed at the first year. So instead of having
actions directed at retention, direct them at first year teaching quality. I
believe this because:
- My attempts to measure academic and social integration (operationalising
Tinto's ideas) have largely shown no problem with social integration at GU.
- Work in the UK by Yorke suggests that the single biggest factor,
especially at relatively high ranking universities, is "academic integration"
or how well a student feels they fit with the course they are doing. For
academics implementing improvements at this, it will feel like improving
teaching quality particularly aspects that give a student enjoyment and
confidence that they are learning successfully at the subject: better real
understanding, and better evidence/feedback to the student that demonstrates
this to them. Often this will involve interaction with other students in
various ways, but where the peer/social processes are subordinated to an
- Jim Boyle's impressive and sustained success at improving retention at
Strathclyde in level 1 Mechanical Engineering (20% dropout reduced to 3%) was
done by focussing on better teaching quality / better achieved learning
experience: it shows it can be done, and done by focussing on teaching and
learning. And it also suggests that focussing on disciplines with relatively
higher dropout (nationally) may pay the biggest dividends. But note it also
implies that the solutions (better teaching and learning) are discipline
- If retention is retained as a goal in itself, then only support
"evidence-based retentioneering" i.e. only measures that have been proved to
have had a real effect in at least one context. In this area, commonsense, or
the judgement of well meaning academics, has a poor record of achievement.
The theory of dropout is interesting, but not well connected with
results. The best approach would be to improve, regardless of benchmarks and
targets of what we "should" do. For this, the best thing is to look at the few
published interventions that have proven effectiveness and apply these, not
invent our own. I have a
list of those I've come across.
And the few proven successes published are also evidence that, regardless of
benchmarks, we could probably do better.
Smith J. & R. A. Naylor (2001)
"Dropping out of university: a statistical analysis of the probability of
withdrawal of UK university students"
Journal of the Royal Statistical Society (Series A) vol.164 pp.389-405
Yorke,M. (1999) Leaving early: undergraduate non-completion in higher
education (London: Taylor and Francis)
Yorke,M. & Longden,B. (2004) (eds. and authors)
Retention and student success in HE (SRHE/OU press)
Web site logical path:
[Top of this page]