Centre for Applied Social Surveys
Short Course
Likelihood Methods for Surveys
Supported by the ESRC
Raymond L Chambers
Alan H Dorfman
Alan H Welsh
9-11 June 2004
University of Southampton
CASS is an ESRC Resource Centre hosted by
the National Centre for Social Research, the University of Southampton and
the University of Surrey
Course Objectives:
To introduce participants to recent developments in the application of likelihood-based statistical methods for the analysis of survey data;
To improve participants' understanding of the statistical theory underlying these methods;
To familiarise course participants with use of this theory by drawing together a variety of applications that make use of its general principles.
Course Content:
The course will cover the following topics:
A comparative overview of different likelihood-based approaches to analysis of complex survey data;
Application to maximum likelihood inference for multinormal and linear regression models based on complex survey data;
Extension to the generalised linear model, particularly the multinomial and logistic specifications, and to the loglinear model for categorical data;
Extension to informative sampling designs, including response-biased sampling and case-control sampling;
Extension to mixed models and multilevel models.
The course will include time for questions and discussion. There will be some exercises in order to reinforce the theory.
Target Audience:
The course is aimed at an audience with a strong statistical background interested in the application of modern likelihood-based methods to the analysis of survey data. Participants will have a good understanding of the 'standard' theory of statistical inference, particularly the theory associated with the use of the likelihood function for inference. They will be interested in learning how the processes involved in data collection, including the use of complex sampling schemes involving stratification, unequal probability sampling and multi-stage sampling, as well as non-response, affect likelihood concepts and what can be done to compensate for them.
Fee:
£400 (includes course materials, lunches and morning and afternoon tea). Because of ESRC support, a number of bursaries are available which reduce the fee to £85 - applicants must be full-time students, UK academic staff or ESRC-funded researchers.
Location:
The course will be held at the University of Southampton. During the course participants can be contacted via the Southampton Statistical Sciences Research Institute (Tel: 023 8059 5376, Fax: 023 8059 3846).
Duration:
The course will begin with coffee and registration at 09.30 am on Wednesday, 9 June and end at 5.00 pm Friday, 11 June 2004.
For further information, please contact:
--
Jane Schofield
Southampton Statistical Sciences Research Institute
University of Southampton
[log in to unmask]
Tel: +44 (0) 2380 595376
|