Planning, Organizing and Funding Longitudinal Studies

John Fox, NHS Health & Social Care Information Center
Jeremy Neathey, Economic & Research Council

Abstract
The UK has a rich heritage of longitudinal data based on individual studies established by scientific pioneers competing for resources in a competitive market place. The value of that heritage is substantially greater now that there is a strategic approach to investment in the research infrastructure that includes a view of what data are needed to address future research which embraces data management and access, as well as well as giving future generations of researchers the skills needed to make the best use of data. With leadership from the Economic and Social Research Council (ESRC) and the Office for National Statistics (ONS) in particular, major funding agencies have come together to develop co-ordinated strategies. Longitudinal research is one of the areas to benefit early from this approach.

Background
This paper draws primarily on our involvement with Research Resources Board of the ESRC and in particular the development of a national strategy for data resources for the social sciences (Peter Elias, 2005) and interim findings of a strategic review of panel and cohort studies (Jean Martin and review team, 2006). We bring diverse perspectives which include those of a researcher, former director of two longitudinal studies, government provider of longitudinal data, health service provider of longitudinal data, and member of committees funding longitudinal research as well as a research council administrator involved in developing the strategic approach.

1. ESRC Strategic Review of Panel and Cohort Studies

Interest in longitudinal data has a long history in the UK. It goes back to the period following the Second World War and interest in pregnancy outcome and maternal and child health and development. It was developed by a number of disciplines including sociology, psychology, social and occupational medicine and drew in others including demography, epidemiology and statistics. The early birth cohort studies in 1946, 1958 and 1970 followed samples of children from birth with periodic interviews, tests and examinations. These studies were developed by leading scientists with substantive research interests and were designed to address their objectives. Funding was sought from a variety of organisations including research councils, foundations and increasingly government and were based on proposals for particular pieces of research.

Each of the research directors complained about the resources that were required to keep these studies going and the limitations that this placed on what they were able to achieve (James Douglas, Mia Kelmer Pringle, Neville Butler video 1970s?). The studies were very expensive to fund. They were multidisciplinary ventures that required strong teamwork. Funding bodies were not sympathetic to investing in maintenance of contact with the study members, data management, providing wider access to data or in development of appropriate analytic methods and improving researcher understanding of and ability to use longitudinal data. These were seen as the responsibility of others whereas these bodies were interested in immediate answers to what they perceived as ‘simple’ questions. There was also an inherent tension in the process of agreeing questions to be included in the survey instruments because of the desire on the part of Government to use these studies to evaluate its policies whereas researchers were more interested in measurements that would have longer-term value.

Despite the challenges, important and valued research came out of these studies and the studies accumulated data that did allow unplanned research on topics of the day. This led to wider recognition both among researchers and the main funding agencies of the importance of longitudinal research. While the value was recognised there were still many barriers to funding around what were perceived as the weaknesses of these studies. Cost, time it took to produce answers, and low response rates (or more correctly loss to follow-up) were those most commonly quoted.

Increased investment went into longitudinal research and new studies were started. The following are current examples of important studies:

Annex 1 provides links to information about the main UK longitudinal studies and how to access them.

In the 1980s and early 1990s a number of reviews were conducted by ESRC (Mildred Blaxter), Medical Research Council (not published?) and by ONS (ref). Each study also came under the microscope before it was funded for major new rounds of data collection – an exercise that often required bringing together consortium funding from different government departments as well as ESRC. These reviews helped to present a clearer and more co-ordinated picture of the importance of these studies and their contribution to current research. They also asked important questions about the limitations and encouraged research funding agencies to take a more collaborative and strategic approach.

Not surprisingly collaboration was not welcomed at first by individual researchers who had established funding for their studies and wanted to retain as much control as they could over the development of their study. They feared that the approach would lead to loss of autonomy and potentially rationalisation. However, the benefits have become more apparent as more funding has gone into these studies over the last few years and the demand for them has grown as more recent graduates working across government and in academia look to use them. It has been evident in ESRC research initiatives that improved access to these resources has also enabled a wider range of researchers to compete in research programmes eg on health inequalities and on families and households.

The ESRC Strategic Review of Panel and Cohort Studies began last year with a workshop in Oxford that brought together a group of leading UK researchers to consider a range of issues including:

It is expected that the review will recommend:

The final report is due in February and will be considered by ESRC’s Research resources Board advised by the UK Longitudinal Studies Centre before its conclusions and recommendations are integrated into the Board’s strategy. The final report can be found at http://www.longviewuk.com/pages/publications.shtml  and a summary is included in Annex 2.

 

2. National Strategy for Data Resources for the Social Sciences

Planning, organising and funding other social science infrastructure resources followed a similar pattern following the Second World War. Initiatives came from individual researchers with little central strategic thinking, planning or encouragement.

The development of new ‘corporate’ social science resources relied on ‘champions’ building up strong support across a wide community to secure support in the funding agencies for larger investments.

In the early days of the Social Science Research Council (ESRC before 1982?) committees represented disciplines which each had their own separate views of what was important, often with a view to preserving or promoting the health of their disciplines. There was a distinction between research project, programme and centre funding and training funding and also some investment in creating a survey archive and in survey and statistical methods.

Over the past few years ESRC has recognised and thought more strategically about its resource centres and programmes. These now include the following:

 

Centres

 

Programmes

These cover a number of functions including activities to:

More information about these activities including links to relevant websites is given in Annex 3.

Two areas outside the longitudinal study field have been significant in forging strong partnerships between ESRC and Government, particularly the Office of Population Censuses and Surveys and latterly ONS. These were the Survey Archive and the development of Census Programme  Both strengthened the data sources by enabling wider access to an increasingly skilled and knowledgeable community that was encouraged to work with the data offices and thereby influenced content, protocols and outputs for these important national resources.

In each of these areas, including longitudinal studies there has now been more than 20 years of collaboration that has resulted in much wider influence on and ownership of what is collected and how it can be used. There is a much stronger empirical research community with a knowledge of, and a history of using these studies.

New challenges however have arisen as there is a growing interest in longitudinal person level, micro-data from new and exciting developments of administrative sources. There is also a desire amongst researchers to join these sources together with other administrative sources, eg across health and education or education and work, and with more traditional survey sources. There has been some discussion about the legal framework and information governance arrangements that are needed to enable such work to proceed. These have highlighted ideas such as a national framework, best practice guidelines, standards and common implementation, effective disclosure control, user licences, safe havens, and data warehouses with licences to link data and create pseudonomised data. It is also recognised that the statistical and research community needs to find ways of influencing these developments, recognising that their primary purposes are around supporting better delivery of public services to individual pupils, patients, clients etc. The route to better data must be through increasing its relevance to front line staff.

One of the challenges facing ESRC in attempting to develop a national strategy for data resources for the social science is the number of organisations that own or have an interest in the main resources. It already has strong bilateral relations with many of the other players but now sees the growing importance of working across these in a more common way.

Towards the end of 2004 ESRC appointed Professor Peter Elias as its Strategic Advisor for Data Resources. His role is to help in the development of potential data resources and to co-ordinate planning between government/agencies and the research councils for the development of data resources. With his support ESRC has now established the UK Data Forum, a group of senior figures including heads of research councils, government departments and private foundations that is chaired by Professor David Rhind, Chair of the Statistics Commission and Vice Chancellor of City University.

“The National Strategy for Data resources for the Social Sciences (the ‘National Data Strategy’) aims to provide a coherent framework for the development and maintenance of a robust data infrastructure, ensuring that relevant and timely data are available to inform and address future research priorities in the social sciences.”

The National Data Strategy is “a plan for the development of data resources by all relevant UK agencies. It involves partnership with and contributions from the major research councils (principally ESRC, MRC and the National Environmental Research Council), all UK government departments, regional and local agencies, the devolved administrations and the major research foundations. However, it is not confined in its scope to the UK. An important feature of the National Data Strategy is that it should help promote social scientific research at the international level, through the development of access to data and research collaborations beyond the UK.”
(Peter Elias 2005 – draft strategy paper to UK Data Forum)

Peter Elias outlines how the strategy will build on views of future research challenges and data needs in particular agreed priority for work on

and assesses how well placed the UK is in relation to:

For the strategy to work it will need to be kept up-to-date. This will need:

As is now normal practice in ESRC the strategy that is developed will be the subject of wide consultation and debate amongst academic and non-academic users of social science research. It will also be subject to periodic review.

2.1. ‘Big Science’ bid for new Household Panel Study
In the past the social sciences have generally seen as the poorer cousin to the ‘hard’ sciences. However the strategic approach that the ESRC has taken over the past 20 years has encouraged it to challenge such perceptions. To demonstrate its growing confidence ESRC has for the first time made a bid for funding under the Office of Science and Technology’s ‘Large Scientific Facilities programme’. Its proposal is to create the “biggest household panel dataset in the world”. Key features of the dataset are:

The study would cost in the region of £50m over about five years. It would provide “a level of granularity that is now essential to address some of the most pressing scientific challenges faced by the UK’s ever more diverse population.”

The outcome of this bid was NOT yet known but the fact that it was one of three short-listed bids demonstrates that ESRC’s confidence was not entirely misplaced. As it turned out, the bid was successful. The final version of the paper will describe more how it is planned to take this survey forward.

 

3. Discussion

Conclusions

 

Annex 1    About the main UK longitudinal studies

This annex provides links that enable the reader to find out more about the main UK longitudinal studies and how to access them. (It still needs work.)

http://www.esds.ac.uk/longitudinal/introduction.asp

http://www.ifs.org.uk/elsa/

http://www.dfes.gov.uk/rsgateway/DB/RRP/u014200/index.shtml

http://datalib.ed.ac.uk/EUDL/facs.html

http://www.dfes.gov.uk/rsgateway/DB/SFR/s000382/index.shtml

http://www.statistics.gov.uk/STATBASE/Source.asp?vlnk=1323

http://www.socialinclusion.ie/EUSurveyonIncomeandLivingConditionsEU-SILC.html

http://www.msoc-mrc.gla.ac.uk/CurrentResearch/20-07/20-07_MAIN.html

http://www.alspac.bris.ac.uk/welcome/index.shtml

 

Annex 2    ESRC Strategic Review of Panel and Cohort Studies

Publications
“STRATEGIC REVIEW OF PANEL AND COHORT STUDIES”
REPORT TO THE RESEARCH RESOURCES BOARD OF THE ECONOMIC AND SOCIAL RESEARCH COUNCIL
by Jean Martin, University of Oxford; John Bynner, Longview; Graham Kalton, Westat (USA); Paul Boyle, St Andrews University; Harvey Goldstein, University of Bristol; Vernon Gayle, Stirling University; Samantha Parsons, Institute of Education; Andrea Piesse, Westat
"APPENDICES TO THE STRATEGIC REVIEW OF PANEL AND COHORT STUDIES”

 

Annex 3    ESRC infrastructure resources: Research Centres, Programmes and policy networks

Links to all resources are to be found at:

http://www.esrcsocietytoday.ac.uk/ESRCInfoCentre/research/resources/

 

Research Centres

Centre for Applied Social Surveys

Summary:

The Centre for Applied Social Surveys (CASS) has been set up with the aim of promoting a fuller understanding and improved application of social survey methods among the UK academic and non-academic communities.
The need for the Centre arises from the view that while there is a clear need to develop reliable quantitative measures to map social and economic change in the UK, there is a shortage of necessary skills to design, manage and analyse large-scale surveys.

To address this problem, the CASS has three main functions:

The series of taught courses was initiated late in 1995. The programme has so far included introductory courses to survey sampling, quantitative survey research, categorical data analysis and handling non-response.

The online service for the question bank was launched in October 1996. It has been designed to be practically useful to the survey researcher by showing how problems of survey measurement can be tackled, how comparability with other surveys can be achieved, and what lessons may be learned from previous attempts to measure key social concepts.

Economic and Social Data Services

The ESDS is a new national data service that is jointly-funded initiative sponsored by the Economic and Social Research Council (ESRC) and the Joint Information Systems Committee (JISC). It is a distributed service, bringing together a number of centres of expertise in data creation, preservation and use:

These centres will work collaboratively to provide preservation, dissemination, user support and training for an extensive range of key economic and social data. ESDS replaces the social science data services offered previously by the UKDA and MIMAS with a single joined-up service offering enhanced support for the secondary use of data across the research, learning and teaching communities. Housed at the UKDA these data collections, numbering over 4000, span both quantitative and qualitative sources across many disciplines and themes.

European Social Survey

The ESS is an academically-driven cross-national social survey, designed as a time series and employing the most rigorous methodologies. It is intended to chart and explain the interaction between Europe's changing institutions, its political and economic structures, and the attitudes, beliefs and behaviour patterns of its diverse populations. More than 22 countries were involved in both the First Round of data collection in 2002, and in the Second Round in 2004. Further Rounds of data collection are planned for 2006 and 2008.

Each national survey is funded by an appropriate national research agency; ESRC funds the UK component of the Survey. Multi-national co-ordination for the first two rounds has been financed by an award from the European Commission's Framework Programme, supplemented by a grant from the European Science Foundation (ESF).

UK Longitudinal Study Centre

Under this award the ULSC is carrying out both national strategic functions and main data production tasks. The ULSC provides intellectual leadership of, and coordination for, longitudinal survey programmes. The award carries forward fifteen years of expertise gained in the collection, documentation and dissemination of the British Household Panel Survey (BHPS) into the production of five further waves of the BHPS. All ULSC activities are underpinned by continuing methodological research on longitudinal data collection and analysis. Results are reported as part of the ULSC's wider dissemination activities. Data collection is disseminated through the ESRC Data Archive. Waves of the BHPS will be deposited annually.

Centre for Longitudinal Studies

The Centre for Longitudinal Studies (CLS) is an ESRC Resource Centre based at the Institute of Education, University of London. CLS houses three of Britain’s internationally renowned cohort studies. These studies involve multiple surveys of large numbers of individuals from birth and throughout their lives. They have collected information on education and employment, families and parenting, physical and mental health, and social attitudes. Because they are longitudinal studies that follow the same groups of people throughout their lives, they show how histories of health, wealth, education, family and employment are interwoven for individuals, vary between them and affect outcomes and achievements in later life. 

The oldest of the three studies is the National Child Development Study, which has followed all the babies born in England, Scotland and Wales in one week in March 1958 (nearly 17,500 infants) from birth to the present day. The British Cohort Study runs along the same lines, taking as its starting point all the births in one week in April 1970 - just under 17,200 babies. The youngest of the three studies is the Millennium Cohort Study - the first survey of this cohort of 18,819 babies was carried out when the babies were nine months old. The intention is to follow this cohort into adulthood like its predecessors. 

Through comparing the different generations in the three cohorts, we can chart social change and start to untangle the reasons behind it. Findings from the studies have contributed to debates and enquiries in a number of policy areas over the last half-century including: education and equality of opportunity; poverty and social exclusion; gender differences in pay and employment; social class differences in health; changing family structures; and anti-social behaviour.
Data from the three studies is available free of charge from the UK Data Archive (www.data-archive.ac.uk). Public access to the studies makes them a scientific and civic resource of immense value to the research, educational and policy communities in the UK and abroad.

National Centre for e-Social Science

NCeSS, the core of the ESRC's e-science strategy, will stimulate use of Grid-enabled computing and data infrastructures in social science research by providing information, training, advice, support and online resources. NCeSS comprises a co-ordinating hub at Manchester University (in collaboration with the UK Data Archive at Essex University) and a network of research nodes across the UK, plus a programme of small projects and agenda setting workshops.

National Centre for Research Methods

The overall mission of the Centre is to provide a strategic focal point for the identification, development and delivery of an integrated national research and training programme aimed at promoting a step change in the quality and range of methodological skills and techniques used by the UK social science community. 
The Centre will be distributed across a number of sites, drawing upon the knowledge and expertise of researchers around the country. It will work closely with ESRC's Research Methods Programme. The Co-ordinating Hub of the Centre is located at the University of Southampton. The Hub has the strategic responsibility for assessing current national provision and identifying future national needs in the development of research methods and training. The Hub will also play a lead role in facilitating the more systematic co-ordination of the Council's existing initiatives aimed at improving the UK skills base in research methods, linking these to other non-ESRC funded activities with similar objectives. The Centre will also consist of 6 Nodes which will work closely with the Hub. The Nodes, which will commence work in April 2005, will be responsible for delivering the Centre's research programme and providing support and dissemination activities across the UK.

The Nodes are:

 International Bibliography of the Social Sciences

IBSS Online is one of the largest and most comprehensive social sciences databases in the world, covering the core disciplines of economics, sociology, politics and anthropology. With its unparalleled international coverage, the database includes monograph records and articles from over 2,400 journals from 60 countries. A total of 100 languages are represented in the database, and of the records appearing in IBSS Online 30 per cent are in languages other than English. IBSS Online coverage dates back to 1951; the database currently contains over 1,700,000 bibliography records, and grows by 100,000 new records each year.

Access to such an extensive collection of publications is essential, as libraries face shrinking budgets and spiralling series costs. IBSS Online provides access to a wealth of new material, even for those libraries with extensive social sciences collections. Since mid-1999 abstracts have been available for up to 50 per cent of new issues of current journals covered. IBSS Online receives support from the ESRC and the Higher Education Funding Council's Joint Information Systems Committee. It is free of charge to all members of the UK higher education community.

 

Programmes

Census Programme

The jointly funded ESRC/JISC Census Programme provides a vital resource for the UK academic community. It holds the only reliable, comprehensive inventory of the UK population and is a key resource for research in social, environmental and geographic planning and medical science in UK universities.
The programme consists of three separate elements:

Research Methods Programme

The principle aim of this Programme is to develop qualitative and quantitative methods within the context of substantive research. It is also geared toward the effective dissemination of good practice through a range of related training activities. Specific objectives include:

Under Phase 1 a total of 321 outline applications were received at the outline stage, with 63 applications invited to submit full proposals. Following full peer review and merit a total of 20 new projects were funded under the first phase of the programme.
As the Commissioning Panel had been disappointed that it was only able to fund 20 projects from the 63 considered under Phase 1 of the programme, it urged the ESRC to supplement the budget to increase the Programme's scope and impact. Additional funding has now been obtained in order to support some of the alpha-rated proposals the Panel was unable to recommend for funding in June 2002. A further 9 projects have now been supported at a cost of approximately £1.35 million. 

The second phase of the programme invited proposals for training activities that will have a direct impact on the methodological skill base of social science research in the UK, a total of 16 applications were received. Following full peer review and merit, a total of 9 new projects were funded under the second phase of the Programme.
One of the major outputs of the RMP to date is the development of a web-based searchable database containing information on research methods short courses. There are currently over 200 courses listed in the database.

British Election Study

The British Election Study is an ongoing programme whose overall aim is to increase understanding of long-term trends in British voting behaviour. The 2001/02 election study had four main elements:

The study design had two supplementary components. The first was a dataset based on Census and other data, which would describe the district-, ward- and constituency-level context for each of the survey respondents. 

Additionally, a content analysis of national press coverage of the election campaign would be conducted to create a 'campaign dataset'. This would be used in relation to opinion changes during the campaign.

Aims of the 2005/06 British Election Study are:

European Social Survey

The ESS is an academically-driven cross-national social survey, designed as a time series and employing the most rigorous methodologies. It is intended to chart and explain the interaction between Europe's changing institutions, its political and economic structures, and the attitudes, beliefs and behaviour patterns of its diverse populations. More than 22 countries were involved in both the First Round of data collection in 2002, and in the Second Round in 2004. Further Rounds of data collection are planned for 2006 and 2008.

Each national survey is funded by an appropriate national research agency; ESRC funds the UK component of the Survey. Multi-national co-ordination for the first two rounds has been financed by an award from the European Commission's Framework Programme, supplemented by a grant from the European Science Foundation (ESF).

Demonstrator Scheme for Qualitative Data Sharing and Research

QUADS is the ESRC Qualitative Archiving and Data Sharing Scheme. The objectives of the Scheme are to:

Scottish Longitudinal Studies Centre

New and not yet on website

Social Science Information Gateway

The Social Science Information Gateway (SOSIG) provides access to relevant high-qualilty Internet resources for social science researchers, academics and practitioners.

Key features include:

Workplace Employment Relations Survey

The 2004 Workplace Employment Relations Survey (WERS 2004) is a national survey of people at work. The survey is jointly sponsored by the Department of Trade and Industry, the Advisory Conciliation and Arbitration Service, the Economic and Social Research Council and the Policy Studies Institute. It follows in the acclaimed footsteps of earlier surveys conducted in 1980, 1984, 1990 and 1998. 

The purpose of each survey in the series has been to provide large-scale, statistically reliable evidence about a broad range of industrial relations and employment practices across almost every sector of the economy in Great Britain. This evidence is collected with the following objectives in mind: to provide a mapping of employment relations practices in workplaces across Great Britain; to monitor changes in those practices over time; to both inform policy development and permit an informed assessment of the effects of public policy, and, to bring about a greater understanding of employment relations as well as the labour market. To that end, the survey collects information from: managers with responsibility for employment relations or personnel matters; trade union or employee representatives; and employees themselves. The information that is provided by respondents to the survey will be used to publish a report that will inform policy-makers and practitioners who work in the field of employment relations. The report will also aid public debate about the nature of work and workplace relations in Britain. 

 

Policy Networks

UK Centre for Evidence Based Policy

The network consists of a national coordinating centre supported by seven network subcentres that will form bridges into available expertise, results, data and information relating to specific major policy issues.

The centre will foster the exchange of research-based evidence between public policy researchers and practitioners, and accelerate the development of methods and capacity for appraising and summarising the results of research relevant to policy and practice across the social science disciplines integrated into the network.

Centre for Economic Policy Research

CEPR provides networking, dissemination and support services to the UK academic community. It also promotes independent objective analysis and public discussion of open economies and the relations among them.

Télécharger le document
Session 4 : Planning, Organising and Funding Longitudinal Studies
Session 7 Potential New Canadian Surveys
Click on the PFD icon above to download the entire paper