Web site logical path: [www.psy.gla.ac.uk] [~steve] [localed] [this page]
Submitted to the 2003 Galway conference of the information and computer sciences LTSN.
ToDo: broken URL to Rebecca's home page, and page on her work; Quintin's exam page.
As in many universities, there is an undesirably high failure rate in the
introductory computer science courses at Glasgow, particularly associated with
programming. A succession of projects here have searched for causes and
predictive markers for failure in this course, partly by testing for
correlations between eventual outcomes and numerous candidate factors such as
entry criteria, study habits and student integration.
The Grumps project offers software that can also be used for
this search. The software records generic usage data (e.g. mouse activity and
typing) and has been deployed in the first year programming environment.
Recording generic data allows an iterative approach to data-mining
investigations in a higher education environment where researchers are limited
to yearly collection. Current investigations concentrate on predicting levels
of success in programming by looking at student persistence, typing and set-up
speeds. This non-invasive approach has uncovered a correlation between
application switching and the students' eventual grade.
Other work is investigating deeper cognitive factors from an
educational and psychological standpoint, exploring links to areas such as
mathematics and language learning.
A third strand of our work seeks ways to increase
student engagement. We have been developing the use of
infra-red handsets in large lectures (cf. Who Wants To Be A
Millionaire?) to promote interactivity and hence engagement. We are extending
the software to support a larger range of interaction techniques, as well as
to make response data available to lecturers and students outside the lecture.
Additionally, a Peer Assisted Learning (PAL) scheme was
introduced in 2002, in which second year undergraduates fill the role of
facilitator and mentor at meetings of first year students. A particular aim is
to support the transition from school-style study habits to those required in
the first year at university, increasingly important as university
participation widens. The first year of this scheme has been evaluated and a
report will shortly be available.
Two further initiatives are of note. Formal
laboratory examinations are a way to sidestep the problems of
plagiarism in continually-assessed coursework exercises. We have developed an
exam that is reliable even when the class size is much greater than the number
of laboratory machines. It complements a traditional written exam by
thoroughly assessing syntactic rigour and debugging skills in the usual
The Lab Support System (LSS) is a
portal to student records, available to both staff via wireless PDAs and
students on laboratory machines. It both mediates the queue of help requests
(replacing students holding up their hands) and provides the circulating tutors
with instant reminders about the students from their records.
Web site logical path:
[Top of this page]