Survey Methods and Analysis Group (SMAG)
The Survey Methods and Analysis Group (SMAG) is a research group in the Cathie Marsh Institute of Social Research (CMIST) led by Professor Natalie Shlomo. The aim of the group is to provide a focal point for researchers in the School of Social Sciences on the development and application of survey methodology and the analysis of complex survey data.
More specifically, they focus on topics such as survey design and estimation, compensating for non-response, analysis of non-response bias using quality indicators, small area estimation, statistical data editing, assessing disclosure risk in statistical data, record linkage and more. The group is composed of members of the Social Statistics Discipline and the UK Data Service.
Events and activities
The Survey Methods and Analysis Group and Statistical Modelling Group have arranged for two experts to visit us in CMIST to present their work on the topic of compensating for missing data in longitudinal surveys. The event will be held on 08/05/2017.
The event is as follows:
George Ploubidis, UCL: 3:00-3:45
Coffee Break: 3:45 – 4:15
Peter Lugtig, University of Utrecht, the Netherlands: 4:15 – 5:00
SMAG Methodology Seminar 3 – 01/03/2016
Gabi Durrant from the University of Southampton spoke on ‘Assessing risk of nonresponse bias and dataset representativeness during survey data collection’.
SMAG Methodology Seminar 2 – 15/12/2015
Nikos Tzavidis from the University of Southampton spoke on ‘Domain Prediction of Complex Indicators – Model-based Methods and Robustness’.
4th International Workshop: Advances in Adaptive and Responsive Survey Design at the University of Manchester - 9-10/11/2015
There were 40 participants to the international workshop from the following countries: United States, Italy, the Netherlands, United Kingdom, Germany, Canada, Norway, Sweden and Australia. The participants included distinguished speakers: Prof. Rod Little and Prof. Jim Lepkowski from the University of Michigan, Prof. Carl Sarndal from Stockholm University, Prof. Roger Tourangeau from Westat, United States, Prof. Peter Lynn from the University of Essex, UK, Peter Miller from the US Census Bureau and Prof. David Steel from the University of Wollongong, Australia.
Careers in Statistics – 04/03/15
Targeted at Biostatisticians this talk was led by Professor Natalie Shlomo. Talking about careers in government statistics, Natalie drew on experiences to brief MSc and PhD students on the complexities and rigours of such a role.
SMAG Methodology Seminar 1 – 22/01/2015
Ton De Waal from Statistics Netherlands and the University of Tilburg spoke on ‘Edit Restrictions in Statistical Data Editing and Imputation’. There were collaborative efforts by the three main statistics groups at the University of Manchester: Social Statistics, Biostatistics and Mathematical Statistics to publicise and attend seminars and events. There was a good turnout for the methodology seminar with over 25 participants.
Showcasing Research in Survey Methods and Statistical Modelling – 03/11/2014
Over 60 participants attended the CMIST kick-off event, which sponsored by the Survey Methods and Analysis and Statistical Modelling Groups. Harvey Goldstein spoke on ‘Linkage is Really a Missing Data Problem: a Bayesian Approach’; Natalie Shlomo spoke on ‘Statistical Disclosure Control Methods and Measurement Error’; David Steel spoke on ‘Efficient Survey Design using Split Questionnaires’; Mark Tranmer spoke on ‘Multiple Membership Multiple Classification Models for Multilevel Network Dependencies, with an Application to Heath Data’.
- Bayesian Adaptive Survey Designs
The proposed network funded by The Leverhulme Trust gathers researchers from academia and national statistical offices and gives a strong impetus to theory development and practical implementation of adaptive survey designs.
Adaptive survey designs differentiate survey design features for different population subgroups based on auxiliary data about the sample obtained from frame data, registry data or paradata. The development of a Bayesian framework will allow the learning and constantly updating of key input parameters to these designs.
- Data Without Boundaries
The Data without Boundaries – DwB – project exists to support equal and easy access to official microdata for the European Research Area, within a structured framework where responsibilities and liability are equally shared.
The InGRID project is funded by the European Union’s Seventh Framework Programme for Research, Technological Development and Demonstration under Grant Agreement No 312691 and involves 17 European partners. Referring to the EU2020-ambition of Inclusive Growth, the general objectives of InGrid – Inclusive Growth Research Infrastructure Diffusion – are to integrate and to innovate existing, but distributed European social sciences research infrastructures on ‘Poverty and Living Conditions’ and ‘Working Conditions and Vulnerability’ by providing transnational data access, organising mutual knowledge exchange activities and improving methods and tools for comparative research.
- Accounting for informative item nonresponse in biomarkers collected in longitudinal surveys (WP3)
The WP3 work package of the National Centre for Research Methods (NCRM) aims to research and develop methods for compensating for non-response in the collection of biomarkers in longitudinal surveys. Non-response can arise from attrition at the interview or nurse visit stage as well as at the stage when blood is collected. Each of the typologies of non-response are driven by different non-response mechanisms and may be informative. Auxiliary information that can be used to compensate for the missing data can cause other non-sampling errors, such as linkage error, that needs to be accounted for when undertaking biosocial research.
- Adrian Byrne
- Georgia Chatzi
- Jose Cuituncoronado
- Angelo Moretti
- Jennifer Taub
- Sixten Thestrup