Edinburgh RSS local group
Meetings 2010



Tuesday 2nd February 2010
David Firth
Getting it right on election night
The Melting Pot, 5 Rose Street, Edinburgh EH2 2PR
Meeting starts 6.30pm, tea and coffee from 6pm
* PLEASE NOTE THAT THE START TIME IS LATER THAN ORIGINALLY ADVERTISED *
Download poster (PDF file)

Abstract
On the day of the UK General Election in May 2005, broadcasters were able to predict the Labour majority of 66 seats, even before any votes had been counted. In this talk some of the statistical methods behind that prediction will be described.

Note
This meeting will be followed by our Annual General Meeting, where we confirm membership for the committee for next year. If anyone would be interested in getting involved in the local group committee, please get in touch with Mairi, Cecilia or any of the other local group committee members.


Tuesday 2nd March 2010
Gillian Lancaster, Lancaster University
Opportunities and innovations in postgraduate statistical education
Sydney Smith Lecture Theatre, University of Edinburgh Medical School, Edinburgh EH8 9AG
***PLEASE NOTE THAT THIS MEETING IS NOT AT OUR USUAL VENUE***
Meeting starts 6pm, tea and coffee from 5.30pm

Directions to venue

Abstract
In 2005, Lancaster University was awarded a large HEFCE grant to fund a Centre of Excellence in Teaching and Learning (CETL). The Lancaster Postgraduate Statistics Centre has the core aim of promoting excellence in teaching statistics to postgraduates across a range of disciplines. For those wanting to continue in higher education, we have two MSc programmes and a PhD programme. For those in employment, wishing to improve their statistical skills, we hold short professional development courses, introductory workshops and masterclasses. This talk will highlight the work of the centre, including the range of training opportunities available for enhancing quantitative skills.


Tuesday 27th April 2010
Adrian Bowman (University of Glasgow)
Surfaces, shape and anatomy
The Melting Pot, 5 Rose Street, Edinburgh EH2 2PR
Meeting starts 6.30pm, tea and coffee from 6pm

Abstract

Three-dimensional surface imaging, through laser-scanning or stereo-photogrammetry, provides high-resolution data defining the shape of objects. In an anatomical setting this can provide invaluable quantitative information, for example on the success of surgery. Two particular applications are in the success of breast reconstruction and in facial surgery following conditions such as cleft lip and palate. An initial challenge is to extract suitable information from these images, to characterise the surface shape in an informative manner. Landmarks are traditionally used to good effect but these clearly do not adequately represent the very much richer information present in each digitised images.Curves with clear anatomical meaning provide a good compromise between informative representations of shape and simplicity of structure. Some of the issues involved in analysing data of this type will be discussed and illustrated. Modelling issues include the measurement of asymmetry and longitudinal patterns of growth.


This page is maintained by Adam Butler.


Tuesday 19th October 2010, 18:00-19:30
The Scottish Longitudinal Study
Joint meeting with RSS Social Statistics section
The Melting Pot, 5 Rose Street, Edinburgh EH2 2PR

An introduction to the Scottish Longitudinal Study: Chris Dibben is Director of the Longitudinal Studies Centre - Scotland and a lecturer in Geography at the University of St Andrews.

The Scottish Longitudinal Study (SLS) is a large-scale linkage study which has been created by using data available from current Scottish administrative and statistical sources. These include Census data, Vital Events data (births, deaths, marriages), Educational data, National Health Service Central Register (NHSCR) data (migration in or out of Scotland) and NHS data (cancer registrations and hospital admissions). It therefore covers a wide range of variables covering cultural, demographic, economic, health, housing and social issues.

Teenage Parents in Scotland 1991 to 2001. Evidence from the Scottish Longitudinal Study:Gillian Raab is emeritus professor of Applied Statistics at Edinburgh Napier University.

Policy initiatives in relation to teenage pregnancies tend to focus on factors that affect rates of conception. There has been less focus on the social, economic and health outcomes for young parents. This information is not readily available from routine sources. The Scottish Longitudinal Study (SLS) allows us to identify young people who become young parents between the 1991 census and the 2001 census and relate this to their health and social status at 2001.

What is the impact of selective migration on the widening mortality gap in Greater Glasgow?: Frank Popham is a research fellow at the University of St Andrews.

The mortality gap between the least and most deprived areas in Greater Glasgow has widened in recent years. Over the same period Greater Glasgow's most deprived areas have seen a significant loss of population, and it has been suggested, therefore, that the widening mortality gap could be, in part, due to internal migration of healthier and wealthier individuals away from these areas rather than a relative worsening of health per se in them. We explore whether this is the case using linked census data from the Scottish Longitudinal Study.


Tuesday 9th November 2010
Chris Robertson (Strathclyde University)
Statistical Issues in Surveillance of H1N1v Influenza in Scotland
The Melting Pot, 5 Rose Street, Edinburgh EH2 2PR
Meeting starts 6.30pm, tea and coffee from 6pm

Abstract
This talk describes the network of surveillance systems used in Scotland to monitor the spread and severity of H1N1v Influenza during the 2009/10 season. Most of the surveillance streams had a statistical exception reporting system imbedded within them and the statistical characteristics of such systems will be described. The most important data streams were the GP consultation rates, the daily reported number of deaths and calls to the NHS24 . Each system generated its own set of statistical problems ranging from reporting delays and how to adjust for them, reporting frequency issues and how to adjust for these and the influence of increased awareness. The talk will focus on the statistical methods employed, the outputs used and the influence of the statistical analysis on the weekly surveillance reports.


Tuesday 14th December 2010
Professor Alexander McNeil (Heriot-Watt University)
Statistical Models for Portfolio Credit Risk
ICMS, 15 South College Street, Edinburgh EH8 9AA
*** PLEASE NOTE THE CHANGE OF VENUE *** Meeting starts 6.30pm, tea and coffee from 6pm

Abstract
We discuss models of portfolio credit risk where the events of interest are the defaults of corporates or transitions in credit ratings. Arguably the most important feature of these models is that they should incorporate a mechanism for modelling the dependence between events due to common exposure to adverse economic factors. The assumptions about dependence determine the shape of the portfolio loss distribution derived from credit risk models and have a major impact on risk capital calculations. In particular we consider models that have the cross-sectional structure of generalized linear mixed models (GLMMs) and the time series structure of hidden Markov models. Due to their complexity they are estimated by computational Bayesian means using Markov chain Monte Carlo. Examples of models fitted to Standard & Poors default and migration data will be given.