THREE PhD STUDENTSHIPS IN STATISTICS
Applications are invited for THREE three-year full-time PhD studentships
at the Open University, Milton Keynes, starting 1 October 2003. The
projects are as follows (for one of the studentships there is a choice of
two topics); details are given at the end of this message.
1. Validation and extension of the self-controlled case series method
This is an EPSRC CASE studentship with additional funding from
GlaxoSmithKline. Supervisors: Dr Paddy Farrington (Open University) and Dr
Thomas Verstraeten (GSK).
2. Quantifying expert opinion as a probability distribution
This studentship is funded by the National Health Service Research
Methodology Programme. Supervisor: Professor Paul Garthwaite.
3. Either: Fitting new families of distributions to data, or: Sensitivity
analysis and robustness
This is an Open University studentship. Supervisor: Professor Chris Jones
(fitting distributions) or Professor Frank Critchley (sensitivity analysis).
Applicants should have at least a 2(i) honours degree or a recognised
postgraduate qualification containing a substantial element of statistics.
The studentships will be based at the main Open University campus in Milton
Keynes. Although the Open University differs from other universities in
that we don't have undergraduate students on site, we are in other ways
similar to other universities. In particular, we have a very strong and
active Statistics Department. For more information about the Department,
see our website at http://mcs.open.ac.uk/Statistics.
Full tuition fees and research expenses will be paid, and a maintenance
grant of 9,000GBP in 2003/4 rising to 10,500 GBP in 2004/5 and 12,000GBP in
2005/6 will be payable . The EPSRC CASE student will receive an additional
4000GBP per annum from GSK.
The EPSRC and NHS studentships are fully funded for students normally
resident in the UK (tuition fees and research expenses, but not the
maintenance grant, are payable to students normally resident in other EU
countries). For the Open University studentship there are no nationality or
residency restrictions.
For further information please contact Paddy Farrington at:
[log in to unmask]
To obtain an application form and the Open University's Research Degree
Prospectus, please email [log in to unmask], or phone +44 (0)1908
654161, or write to:
Statistics Studentships
The Recruitment Secretary
Faculty of Mathematics and Computing
Walton Hall
The Open University
Milton Keynes MK7 6AA.
You can also download the prospectus and application form from
http://www.open.ac.uk/research-degrees/.
Where the application form asks for details of your research topic, simply
state which (one or several) of the projects you are most interested in,
and why. Completed application forms should be returned by 30 June to the
Recruitment Secretary at the address above (or emailed to mcs-
[log in to unmask]).
Equal Opportunity is University Policy.
PROJECT SUMMARIES
1. "Validation and extension of the self-controlled case series method"
The self-controlled case series method is a novel and powerful method for
estimating relative risks of events following transient exposures, and has
been widely used to investigate adverse reactions to vaccines and other
drugs. The project will involve a detailed investigation of the performance
of this method compared to cohort and case-control studies, using
simulations and analytic approaches. The student will also investigate ways
of weakening the assumptions upon which the method is based. Further
aspects of the project might include: improving the design of case-series
studies, including sample size formulae, and extensions of the method to
applications other than drug safety. Over the three years, the successful
applicant will spend a minimum of three months at GSK Biologicals in
Rixensart, Belgium. Travel and accommodation expenses will be paid.
2. "Quantifying expert opinion as a probability distribution"
Explicit use of Bayesian methods requires that the probability distribution
describing an individual's prior beliefs about the true value of the model
parameters, is combined with the full likelihood for the parameters arising
from the data. Tractable methods for combining these sources of information
are now available for increasingly complicated and realistic models.
Quantifying individuals' beliefs as a probability distribution, however, is
difficult both conceptually and practically, and remains under-researched.
The proposed studentship will form part of a larger project that is
conducting a review of, and primary research into, approaches to the
elicitation of prior probabilities. The plan is that the student will
contribute to the review, conduct experiments in which medical
practitioners quantify their opinions using a variety of techniques, and
develop practical methods for eliciting prior distributions.
3(a). "Fitting new families of distributions to data"
New families of distributions have recently been developed with a variety
of skewness and tailweight properties for use in statistics. The project
will emphasize issues in the practical fitting of these distributions to
data using maximum likelihood. Initially, simulation studies will assess
the usefulness of asymptotic formulae and issues such as parameterisation
in this context. Further developments will involve one or more of the
following: use as residual distributions in regression (and perhaps in time
series, where applications are plentiful in finance and economics);
distribution theory, including further new distributions if necessary;
multivariate versions. An important outcome will be the provision of user-
friendly software in a popular language such as, for example, R.
3(b). "Sensitivity analysis and robustness".
There has been a great deal of recent research activity in these two
related areas. Sensitivity analysis asks: 'For a given statistical
analysis, how well have the assumptions on which it is based been met?'
and 'What are the consequences of departures from them?'. On the other
hand, robust statistical procedures are modifications of classical
inferential procedures, designed to be relatively unaffected by departures
from their underlying assumptions. It is envisaged that the project will be
specified within these two areas according, in part, to the background and
enthusiasm of the student.
|
