Last month CLOSER joined 800 fellow delegates for the seventh conference of the European Survey Research Association (ESRA).
The week-long event took place in Lisbon, Portugal between July, 17 – 21. The marathon conference was hosted by the University of Lisbon’s School of Economics and Management, the Centre for Research in Social Sciences and Management, and the Institute for Social Sciences.
Keynote speakers
Professor Dr Edith de Leeuw (Utrecht University) kicked off the conference with her keynote presentation, “Mixed Mode: Past, Present and the Future”. Prof de Leeuw began with a detailed history of mixed mode, noting that one of the first times it was mentioned was in 1988 by Dillman and Tarnai in a methodological monography. Moving into the present day, de Leeuw suggested there are three important reasons for using a mixed-mode survey design: improving coverage, increasing response rates and reducing costs.
However, there are disadvantages to implementing a mixed-mode design which include an increased administrative and logistic burden, and the potential for bias. To help reduce this risk, de Leeuw suggested survey designers need to aim for equivalent questionnaires across the different modes. For instance, the option to select ‘Do not know’ as a response is not explicitly given during interviews but is often included in web-based surveys. By creating equivalency across different modes, the risk of bias is reduced.
Prof de Leeuw also spoke about the challenges brought about by new technologies, specifically mobile devices. The rising use of internet-ready devices mean web-based surveys should be moving to multi-device oriented concepts.
Presented by Beth-Ellen Pennell (University of Michigan), the second keynote presentation, “Trends and developments in multinational, multiregional and multicultural (3MC) surveys” took place on Thursday morning. Whilst discussing the trends and new developments within 3MC surveys, Pennell also highlighted the challenging dimensions of this research through a Total Survey Error (TSE) model adapted for 3MC studies.
During her talk, Pennell also focused on the opportunities that affordable technologies are bringing to low resource settings. Data collection technologies usually used in high income countries are now being adapted in order to be used in new contexts. For instance, the use of global positioning systems (GPS), aerial photography and fingerprinting are all being implemented in lower income countries.
Papers featuring CLOSER work
Three representatives for CLOSER, Andy Boyd (University of Bristol), Carli Lessof (University of Southampton), and Will Poynter (CLOSER), presented at the conference.
‘Understanding young people’s views about consenting to data linkage: findings from the PEARL qualitative study’, presented by Andy Boyd
Established with the aim of obtaining consent for data linkage between ALSPAC participants and administrative data source, the Project to Enhance ALSPAC through Record Linkage (PEARL) also collected qualitative research to examine participants’ views about data linkage.
Researchers used digitally recorded interviews with 55 young people, aged between 17 and 19 years old. The participants were presented with four scenarios of linking different data types and sources and were asked to consider whether or not consent needed to be requested.
The project found that scenarios relating to teenage pregnancy and mental health created a sense of unease about individuals being stigmatised or blamed. In contrast scenarios relating to heart disease and asthma tended to be seen as having a clearer purpose and health outcome, suggesting a preference for research with tangible health benefits.
Accommodating these views within a governance framework that is acceptable to a majority of the public is challenging. Pragmatic, imaginative and flexible approaches are needed if research using data linkage is to successfully realise its potential for public good without undermining public trust in the research process. Robust evidence such as these are able to inform national strategies to improve the understanding and acceptability of data linkage.
‘Testing for differences in measurement devices: findings from a randomised trial to compare measures of physiological function and physical performance’, presented by Carli Lessof
We are immensely proud that CLOSER has funded Carli’s important work examining the reliability of using different devices to collect biological measures. The reliability of measurement devices is integral to the reliability of the data produced. Carli’s work focused on assessing the measurements made by two spirometers, four dynamometers and two sphygmomanometers in a randomised cross-over trial. The sampled consisted of 118 healthy individuals aged 45-74.
On the whole, a comparison between devices showed no need for adjustments apart from a few exceptions. A possible difference was visible between the measurements of the two hydraulic dynamometers assess, while there was a significant between the measurements taken by the electronic and hydraulic dynamometers. Finally, the difference found between the two spirometers access may be too great for adjustment.
‘CLOSER Repository: Modernising Longitudinal Study Management’, presented by Will Poynter
Our technical manager gave a thought provoking presentation during the ‘Putting the Data in the Driving Seat’ session. Will first outlined the difficulties of data persistence using an example of data loss by NASA, the most well-funded research organisation in the World. CLOSER is pulling together documentation from over 70 years of data collection, while also considering how data can be documented and protected going forward for decades and perhaps even centuries.
Will gave a quick round up on the work already completed to enhance the data and metadata of our constituent studies. But he did not stop with what has been done and spent the second half of his presentation explaining the many options and avenues now available to studies and researchers. Although this started relatively simply, like automated PDF generation, it quickly moved on to move challenging area such as using deep learning to clean data or read old questionnaires automatically.
Although some of these things seemed impossible five years ago when CLOSER started, Will’s presentation lead to a lot of discussion afterwards about the potential of these ideas and proactive ideas about how the technical/data communities can come together to create some of these solutions. It seems that technical teams from across the World are now seeing opportunities that very recently seemed impossible.
Click here for the slides.