17.5 hours per week for 18 months.

We are seeking a motivated researcher in social statistics or demography with excellent analysis skills to work on an Economic and Social Research Council funded project on the effects of social interaction and participation on cognitive functioning and on service use.The post will be based in the Faculty of Health Sciences, University of Southampton (www.southampton.ac.uk/healthsciences).

The project is part of the major ESRC award at LSE on 'Comprehensive approach to modelling outcome and cost impacts of interventions for dementia.'  This is collaboration between UK academic institutions: University of Southampton, London School of Economics, the Universities of Sussex, and Newcastle, and will involve occasional travel to meetings with partners at LSE.

The aims of our project are to investigate, using  existing longitudinal datasets,  i) whether long-term social interaction and participation, by type, have direct, or indirect, protective effects on cognitive functioning in early and later old age and ii) the effect of accumulated social support networks on formal help seeking and receipt of services among those with cognitive impairment.

The successful candidate will perform analyses of existing secondary data: 1) a life course dataset [British birth cohort – the National Child development Study, NCDS – waves 0-8], and 2) a later life cohort [English Longitudinal Survey of Ageing, ELSA – waves 0-5]. The datasets are accessible via the UK Data Archive upon user registration (www.ukdataservice.ac.uk). The post will involve responsibility for merging each wave of NCDS dataset, managing and analysing data(descriptive statistics to multi-variable models of temporal sequence (cross-lagged temporal regressions and structural equation modelling).

The ideal candidate will have excellent statistical and research skills, including use of advanced multi-variable models, and interpreting statistical output. The successful candidate will have experience of preparing and analysing large and complex longitudinal datasets using advanced statistical methods and models, be expected to contribute to the presentation of research findings at meetings and conferences, and publications in articles submitted to top ranking journals. Ability to work to deadlines is essential.

Further details of the datasets and analyses to be conducted:

NCDS and ELSA used the same measures of cognitive functioning at key waves (ages 16 and 50 NCDS; ages 50+: Wave 1-5 ELSA):  word list recall, animal naming, letter cancellation and delayed word list recall. ELSA derived an index by combining index scores (to be re- computed with care as ELSA omitted one indicator in later waves). Both datasets, at various waves, included measures of social networks (network size, type, quality); support from the network (instrumental and emotional), interaction (contacts with social network members), participation (engagement in social and leisure activities, including physical activity; civic engagement), service use. Both also include measures of childhood circumstances (collected retrospectively in ELSA). 

The analyses will include: i) Cognitive functioning at age 50 (NCDS), trajectories of cognitive function after age 50 (ELSA), and covariates; ii)Use of formal services (by cognitive function), and co-variates. Covariates for both aims, (time varying where appropriate), frequency of social interaction - contacts by type and frequency; indicators of social participation and activities,); socio-economic and psycho-social variables; health related behaviours; childhood circumstances. All analyses will adjust for the effects of socio-demographic, socio-economic and other relevant variables (including level of education, SES).

Planned statistical methods include multi-variable models of temporal sequence (cross-lagged temporal regressions) to account for any effects of changes in the predictor variables on cognitive functioning; Structural equation modelling to examine mediators. It is intended to deal with missing data by using multiple imputation techniques. The results will be used to inform modelling by our LSE partners of projected dementia, service use costs, and costs of preventive levels of social engagement and participation. Our LSE collaborators will also undertake Growth curve models using the datasets [http://pathways.lshtm.ac.uk/].

The closing date for this post is Friday 28 May 2014; interview date 12 June 2014.

Informal enquiries may be made via email to Professor Ann Bowling by email [log in to unmask]
Application Procedure:

Please submit your online application through www.jobs.soton.ac.uk. If you need any assistance with this, please contact Carol Read + 44 (0) 23 8059 3649. Please quote reference number 406714BN on all correspondence. 

 

You may leave the list at any time by sending the command

SIGNOFF allstat

to [log in to unmask], leaving the subject line blank.