Survey Futures: ECRs are changing the game

Forget what you think you know about surveys. This isn't your dusty clipboard and door-to-door interviews. The 1st Survey Futures Early Career Researchers Conference at the University of Essex proved that early-career researchers (ECRs) are not just keeping surveys alive – they're reinventing them. From tackling tricky within-household selection to harnessing the power of video interviews, ECRs are pushing the boundaries of survey methodology and ensuring its continued relevance in the 21st century. The conference, featuring twelve presentations, a panel discussion, and a keynote address, provided a valuable forum for ECRs to engage with the evolving landscape of survey methodology. Here's a look at some of the game-changing research presented at the conference.

Viktor Sladka, Senior Research Officer, University of Essex, opened the conference with an exploration of targeted procedures in self-completion surveys, focusing on improving efficiency and cost-effectiveness. Viktor's work examined tailoring survey approaches to specific subgroups, a practice less common in online surveys compared to traditional face-to-face interviews, offering potential for optimisation in digital data collection. This research could lead to more efficient and cost-effective online surveys, allowing researchers to gather data from diverse populations with greater precision.

Nhlanhla Ndebele, PhD Researcher at City, University of London, addressed the complexities of within-household selection methods for surveys without field interviewers. This is a significant challenge when having to rely on address-based sampling frames. Nhlanhla's research aims to develop best-practice guidelines for this crucial area, which could significantly improve the representativeness of self-completion surveys relying on address-based sampling.

Cristian Domarchi, Research Fellow, University of Southampton, analysed recruitment practices in UK social surveys conducted without field interviewers, seeking to identify optimal strategies for survey design and achieving representative samples—a key concern given the increasing reliance on online data collection. Cristian's findings could inform best practices for online survey recruitment, leading to more representative and reliable data.

Nathan Reece, Survey Project Manager, drawing on experience with the European Social Survey, presented research comparing within-household selection methods in self-administered push-to-web surveys, evaluating the impact of a cost-effective non-random method against a random approach on population representation, estimation precision, compliance, and costs. This research will offer practical guidance for survey designers seeking to balance cost-effectiveness with data quality.

Turning to the influence of survey modes, Marc Asensio-Manjon, Research Fellow, University College London, assessed the impact of video interviewing on survey measurement and data quality. Marc's experiment compared video, web, and in-person survey modes to understand the similarities and differences in resulting data. This research helps to understand the potential of video interviewing as a survey mode and its impact on data quality.

Jim Vine, Senior Research Officer, University of Essex, investigated how consent to data linkage varies depending on survey mode, specifically examining the effectiveness of re-asking non-consenters in web mode compared to face-to-face, using data from a major UK household panel survey. Jim's findings could inform strategies for improving data linkage rates in different survey modes.

Jasmine Mitchell, PhD student, University of Essex, examined how household panel members respond to repeated requests for diverse additional data, analysing participation in various tasks (e.g., mobile app usage, bio-measure provision, data linkage agreement) and exploring potential negative impacts on panel retention. Understanding these dynamics is crucial for maintaining participant engagement in long-term panel studies.

Richard Bull, Researcher, National Centre for Social Research (NatCen), discussed strategies for reducing mode effects in surveys through effective questionnaire design, providing an update from the Survey Futures programme and offering guidance on assessing and mitigating measurement non-equivalence as we transition to online and mixed-mode surveys. Richard's work provides practical tools for researchers navigating the complexities of mixed-mode survey design.

Following lunch, a stimulating panel discussion, moderated by Thomas O'Toole, Postdoctoral Research Associate, University of Manchester, explored the crucial role of good practice guides in survey research. The panel, comprising Alun Humphrey (NatCen), Andrew Phelps (ONS), and Eileen Irvin (Ipsos), included representatives from leading UK social research organisations who will be  amongst the prime users of these guides, and offered valuable perspectives on their application and importance, exploring the characteristics of effective guides. This practitioner-focused discussion provided a nuanced understanding of their vital role in ensuring high-quality social research.

The keynote address, "Do you think surveys have a bright future? I do," delivered by Sarah Henry, Director of Methodology and Quality at the Office for National Statistics (ONS), presented a compelling case for the continued relevance of surveys in research. Sarah addressed the perception of surveys becoming obsolete due to publicly available data and concerns about AI rendering them unnecessary. While acknowledging these challenges, she emphasised the unique strengths of surveys and their enduring importance. Drawing on the ONS' experience with the Labour Force Survey (LFS), Sarah highlighted the difficulties encountered when attempting to replace or supplement it with "free" public data, often containing hidden biases and lacking the depth and precision of well-designed surveys. Sarah also reassured the audience that fears of AI obsolescing surveys are likely overstated, emphasising the cyclical nature of enthusiasm for new technologies and the enduring value of established methodologies like surveys, with their capacity to collect highly specific data.

After the keynote, Michalina Siemiatkowska, Senior Research Officer, ONS, discussed optimising "Knock-to-Nudge" data collection for the Transformed Labour Force Survey, describing a project aimed at improving the efficiency and effectiveness of field interviewer visits to non-responding households and encouraging online or telephone participation. This work has the potential to improve response rates and reduce survey costs.

Thomas O'Toole spoke again in the afternoon session, providing a comprehensive overview of integrating survey and non-survey data, covering the rationale, diverse methods and data sources, and the added value derived from combining survey data with administrative, geospatial, and digital trace data. Thomas' presentation highlighted the power of data integration for enriching survey data and gaining deeper insights.

Pablo Cabrera-Alvarez, Survey Methodologist, University of Essex, investigated the need for non-response follow-ups in web surveys of the UK general population, examining the cost-quality trade-offs, particularly focusing on how follow-ups might improve representativeness and potential measurement differences across modes. Pablo's research provides valuable guidance on optimising follow-up strategies in web surveys.

Finally, Jamie Moore, Research Fellow, University of Essex, considered the impact of mixed-mode survey design on the performance of UKHLS COVID-19 Study datasets, evaluating the impact of web and telephone modes and exploring the potential benefits of expanding telephone sampling in mixed-mode designs. Jamie's findings are relevant for understanding the impact of mode effects on data quality in mixed-mode surveys.

In summary, several key themes emerged from the 1st Survey Futures ECR Conference. A strong focus was placed on optimising online and mixed-mode surveys, reflecting the increasing reliance on digital data collection. Researchers explored various strategies for improving efficiency, cost-effectiveness, and representativeness in these contexts. Another recurring theme was the importance of data quality, with presentations addressing issues such as mode effects, within-household selection, and non-response bias. The conference also highlighted the growing interest in integrating survey data with other data sources, such as administrative records and digital trace data, to enhance research insights.

This conference highlighted the dynamism and innovation in survey research. The presentations demonstrated ECRs' commitment to ensuring the continued relevance and effectiveness of surveys by exploring new methodologies and refining existing ones. The conference also provided a valuable opportunity to connect with fellow researchers passionate about the future of surveys. 

The future of surveys appears promising, thanks to the dedication and ingenuity of researchers working at the forefront of survey methodology. The research presented at the conference has significant implications for the future of survey methodology. The innovative approaches and methodological refinements discussed have the potential to improve the quality, efficiency, and relevance of surveys across a wide range of social research applications. By addressing key challenges and exploring new frontiers, these ECRs are contributing to the vitality of survey research, changing the game to ensure the continued relevance of surveys in the 21st century.