Topic A: Establishing patterns of error in household and individual level consumption dataThis project concerns the relationship between a range of consumption patterns and energy consumption levels. The project will review approaches to the estimation of reported 'consumption error'; combine and/or link consumption and other datasets from various sources to produce a set of experimental linked consumption datasets; develop methods to analyse the characteristics and distribution of error in reported consumption of, for example, water and energy.
Topic B: The Residential Mobility of Mental Health Service UsersThe aim of this research is to enhance understanding of the residential mobility of patients receiving NHS mental health treatment. Tracking geo-referenced patient movements via administrative sources, the project will undertake multilevel longitudinal analysis of individual and area variables to predict the numbers of moves, comparing trajectories of people with mental illness with a comparison sample of others treated for conditions that are also likely to give rise to multiple hospital treatment episodes.
Topic C: Multistage cluster sampling using geo-frameIn traditional approaches to multistage sampling design, the clusters are fixed in advance. This PhD concerns investigation of multistage cluster sampling methodology in which geo-referenced information is incorporated into the sampling frame. The project will explore the possibility of using administrative data to perform dynamic clustering in order to minimize the intra-cluster correlation subject to the constraints of cost or work load, and thereby achieve to better trade-off between cost and design efficiency.
Topic D: Combining survey and administrative data for nonresponse investigationThis project aims to combine survey data with administrative data to analyse survey nonresponse behaviour. The project will use multilevel modelling to investigate key determinants of nonresponse and representativity indicators and to develop nonresponse adjustment methods. The project will inform survey practice on how to analyse and adjust for nonresponse using administrative data. The research will improve survey designs including adaptive and responsive survey designs.
Details of how to apply are provided on the Southampton ESRC DTC website at http://www.southampton.ac.uk/esrcdtc. Applicants must apply to the University and submit the additional DTC form downloadable from the website. The closing date for applications is Monday 28th February 2014 and interviews will be arranged with the short-listed applicants soon after this closing date. These studentships are for entry in October 2014.