The Royal Statistical Society (Social Statistics Section) is holding a half
day meeting
Performance Indicators in Higher Education: Statistics of Learning and
Teaching. Tuesday 17th October 2000 from 2:00 until 5.45,
at the RSS, 12 Errol Street, London EC1Y 8LX, tel 020 7638 8998
2.00 Statistical Analysis of Performance Indicators in UK Higher
Education. David Draper and Mark Gittoes (University of Bath)
Performance Indicators (PIs) are outcomes that are thought to be
measures of the quality with which universities are carrying out their
public mandate. Several PIs for UK universities have recently been
published by the Higher Education Funding Council (HEFCE).
Constructing such PIs raises fundamental questions of observational
data analysis, including causality, confounding and standardisation.
We illustrate these statistical issues by analysing data on student
drop-out, using multilevel models and methods for (very) large (very)
sparse contingency tables.
3.00 Data Cleaning: from Rags to Riches. John Thompson and Mark Corver
(HEFCE)
The Higher Education Statistical Agency student data sets do
not provide information relating to course completion. Through a
process of removing duplicates, patching data from other sources, data
correction and matching, a longitudinal dataset has been created so
that student progression through changes of course or institution can
be followed. This data set underpins the HEFCE PIs and has the
potential to answer many key policy questions in Higher Education.
3.45 Tea
4.15 Dropping out of University: a Statistical Analysis of the
Probability of Withdrawal for UK University Students. Jeremy Smith and
Robin Naylor (University of Warwick)
Using individual-level data for an entire cohort of undergraduate students
in
the UK, we fit a probit model to estimate the probability that a
student will 'drop-out' of university. We find evidence to support the
hypothesis that student completion is influenced by the extent of
academic preparedness, and show that social integration at university
is also important. We draw conclusions regarding the public policy of
constructing PIs in this area.
5.00 Discussion 5:45 Close
............................................................................
...........................................
ALL WELCOME MEETING IS FREE
............................................................................
...................................
I will be attending the meeting of the Social Statistics Section on 17
October 2000
Name ......................................................................
Address.....................................................................
............................................................................
.......
Postcode ................................
Telephone
..................................................................
Email
.........................................................................
Please return to Grace Lococo by 12 October at the RSS ([log in to unmask])
NB Registration is not compulsory as this is an open meeting but, in the
event of a change in arrangements, it does help if we have contact details
for people planning to attend
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|