Preventing and Compensating for Missing Data in Biosocial Longitudinal Data
|Dates:||15 November 2018|
|Times:||14:00 - 17:00|
|What is it:||Talk|
|Organiser:||Cathie Marsh Institute for Social Research|
|Who is it for:||University staff, Adults, Alumni, Current University students|
We present on-going research at the University of Manchester on preventing and compensating for nonresponse in biosocial longitudinal data. Fiona Pashazadeh will open the workshop with a systematic literature review based on two large longitudinal surveys in the UK - Understanding Society and the English Longitudinal Study of Ageing (ELSA). Published papers will be categorised according to which, if any, missing data techniques have been used and how the possible impact of the missing data on the resulting analysis has been discussed. Georgia Chatzi will present research examining the association between socio-economic position and the biomarker C-Reactive protein (CRP) over four waves of data from ELSA. A latent growth curve model is presented which includes different imputation procedures and joint modelling to compensate for the missing data with a comparison to the complete case analysis. Following the break, Tina Hannemann will present initial results from a large scale simulation study based on ELSA comparing methods for compensating for missing data under a growth curve model to assess trajectories of the biomarker CRP. Finally, Tarani Chandola will present methodological issues for handling large amounts of attrition for a study based on a longitudinal analysis of refugees’ self-reported health and well-being.
2:00 - 2:30 Refreshments
2:30 - 3:00 Fiona Pashazadeh
What are the current approaches taken to deal with missing biological data due to nonresponse in population representative sample surveys? A systematic review of the literature.
3:00 - 3:30 Georgia Chatzi
Life course socioeconomic position effects on repeated systemic inflammation in older adults: compensating for missing data.
3:30 - 4:00 Break
4:00 - 4:30 Tina Hannemann
Compensating for Non-response in Biosocial Research: Simulation Study from a Longitudinal Data Analysis
4:30 - 5:00 Tarani Chandola
What are the wider social determinants of refugee health and well-being? Some methodological considerations in a longitudinal survey of refugees in the UK
No booking necessary
Travel and Contact Information
Humanities Bridgeford Street