Susan Cassell is the ESRC Communications Manager. She recently visited the seventh ESRC Research Methods Festival, held at the University of Bath.
Here she provides an overview of the three-day event.
What is the latest when it comes to social science research methods?
This was my quest to uncover as I visited my first Research Methods Festival – an inauguration into the fascinating, ground-breaking, albeit at times ‘anoraky’, world of research methods.
This biennial event, established in 2004, is organised by the ESRC-funded National Centre for Research Methods.
It engages social scientists across a wide range of disciplines and sectors at all stages in their research careers.
I thoroughly enjoyed the opportunity to network with researchers at the same time as learning about latest thinking and developments in social science research methods.
Held at the University of Bath for the first time, the Festival had a tremendous buzz, and it was great to be in on the conversations that take place outside of the formal sessions – here, connections are made, new ideas and ways of thinking explored, in a spirit of friendship and co-operation. The beauty of the surroundings (and the intermittent sunshine) helped to make this a truly excellent event.
So, on to my question… The programme offered just what I needed to explore this with more than 50 sessions over three days, covering topics ranging from cohort and longitudinal methods, analysis of complex data sets, and new methodologies for record linkage and use of administrative data for social science research.
There were two key-note addresses on research methodology, both highlighting challenges to the research community and possible responses.
The first by Professor Jane Elliott, CEO of the ESRC, who gave the NCRM Annual Lecture ‘Big Data: Bridging the Qualitative/Quantitative Divide?’
Professor Elliott explored the extent to which the development of machine learning techniques to analyse large bodies of unstructured data (such as text) are blurring the differences between qualitative and quantitative approaches to research in the social sciences.
Three main challenges for social scientists in the era of Big Data were set out:
- The methodological challenge – how we can develop the very best tools to help us interrogate, analyse and understand the vast quantities and varieties of data that now exist
- The challenge of developing insightful research questions and the need to focus energy on the substantive evidence that can be gleaned from the empirical material
- The ethical questions raised by new forms of data and new approaches to analysis.
Professor Elliott concluded that researchers need to be confident about what they bring to the table as social scientists, ‘but also need to be ready to learn from other disciplines to develop and renew our methodological toolkit’.
The second key-note speaker Professor Andrew Gelman,Professor of Statistics and Political Science, and Director of the Applied Statistics Center at Columbia University (USA), spoke about ‘Crimes against Data’, pointing to some recent highly publicised examples where social scientists have used statistical methods to claim ‘statistical significance’, when these claims are implausible and could not be replicated by independent research teams.
The lecture offered a way out of this hole, though he stressed that it required effort and focus on measurement and its close connection to social science models, rather than through any simple recipes of statistical analysis or replication.
Gelman discussed examples of successful research that proceeded by systematically answering many questions at once rather than attempting to estimate different effects in isolation, as is typically done in the ‘junk science’ studies that do not replicate.
The event also show-cased some ESRC-funded data infrastructure investments, including:
- The Administrative Data Research Network (ADRN), which is part of the the ESRC’s Big Data Network.
Dr Kakia Chatsiou of the Administrative Data Service (University of Essex) provided an overview of the administrative data landscape in the UK, what services the ADRN can offer to academic, public and third sector communities, and the innovative research that can be achieved by using the Network. Together with Professor Peter Smith, Director of the Administrative Data Research Centre – England, they gave examples of administrative data research currently underway in the Centres.
Professor Chris Dibben, Director of the Administrative Data Research Centre – Scotland, spoke about administrative data on mothers and siblings and the importance of the data in reconstructing the past to produce new 20th century cohort studies. He also highlighted the Centre’s work on public engagement and on the legal aspects of administrative data.
Dr Dermot O’Reilly, Director of the Administrative Data Research Centre – Northern Ireland, provided insights into the ways death is recorded in longitudinal census data, and the methodological challenges that arise for research.
Professor Ronan Lyons (Swansea University) outlined work carried out at the Administrative Data Research Centre – Wales, including the building of collaborative relationships with stakeholders for the evaluation of policy and interventions.
- CLOSER (Cohort & Longitudinal Studies Enhancement Resources)Professor Alison Park (UCL Institute of Education) compared data across and within longitudinal studies: challenges, opportunities and resources.Professor Rebecca Hardy (UCL MRC Unit for Lifelong Health and Ageing) and Dr Claire Crawford (University of Warwick and Institute for Fiscal Studies) presented key learnings from CLOSER-funded projects that have compared data on BMI and socio-economic status from a number of different UK cohort studies via retrospective harmonisation of key measures. This approach allows the experiences of one cohort to be compared with those of older and younger cohorts, giving us far greater insights into how society is changing.
Dr Bilal Nasim (UCL Institute of Education) described a longitudinal schools database created by CLOSER and the session finished with a presentation about CLOSER Discovery, a resource to help researchers uncover what questions have been asked across different longitudinal studies.
- Understanding Society – the UK Household Longitudinal StudProfessor Michaela Benzeval, Director of Understanding Society, Professor Peter Lynn (ISER, University of Essex) and Dr Tarek Al Baghal (ISER, University of Essex) presented on innovations in methods for longitudinal data collection and analysis.A core principle of Understanding Society is to ensure it is based on the best methodology.Key issues discussed were: moving to mixed mode data collection to save costs while maintaining response levels and data quality; question design and measurement issues in a panel context and using targeted approaches to improve retention.New survey methodologies are trialled by the team and external researchers in the Understanding Society Innovation Panel.
During other sessions at the Research Methods Festival 2016, leading academic experts presented latest innovations in topics such as:
- data visualisation
- diary methods
- time-use diary data
- use of biomarkers in bio-social research.
This year’s Festival had an international flavour, with a large number of delegates and presenters from overseas and an expert panel session on data collection in zones of violence and armed conflict. The panel discussed the potential for more mixed method collaborations in order to overcome the challenges of conducting research in these zones.
An interactive workshop on ‘making the most of the media’ was offered to all delegates, as well as skills development sessions on: ‘writing up your PhD’; how to read and write critically, developing effective research proposals, and capturing research impact.
Thanks especially to the NCRM Festival organising team, who did a fantastic job: Rosalind Edwards, Jonathan Earley, Alexandra Frosch, Rebekah Luff, Eva Nedbalova, Jacqui Thorp and Penny White, ESRC National Centre for Research Methods, University of Southampton.