One of NESSIE’s basic aims is to identify major activities that would aid researchers in carrying out European comparative research. Through its various activities, in particular the organising and hosting of Roundtables on aspects of the research process, the following recommendations for future action have been formulated. These have been extended throughout the life of the NESSIE project. For full reports on the Roundtables, see http://www.nessie-essex.co.uk.
NESSIE and European Research Infrastructure:
The Concept of Research Infrastructure in the Socio-Economic Sciences
- The definition of research infrastructures for the social science needs urgent and well-founded discussion within the social science community.
- There is a need for coordinated exchange of information and for action between all those involved in data production and analysis. Joint initiatives should be organized and followed through as widely as possible.
- Active involvement of Eastern European countries in all activities needs to be further encouraged.
- The collaboration between the European data archives should be used as a model for the development of the wider research data infrastructure.
- There is an urgent need to establish good links at the national level between the research community and the National Statistical Institutes.
- Both the necessity of and the infrastructure needs for the social sciences are accepted, and the NESSIE recommendation and findings seem to fall in line with discussion both within the European Commission and the European Science Foundation Standing Committee for the Social Sciences. From NESSIE discussions, it is clear that the work of NESSIE is considered helpful in the definition and development of the social science infrastructure discourse in Europe and should be continued.
- Infrastructure is clearly high on the European agenda. There are clear indications of needs and organisational proposals which have to be worked out. The OECD data policy declaration, for example, is also endorsed by the European Union and the European Science Foundation.
- Future development must be based on the needs voiced by researchers themselves. EROHS is an idea which must be implemented, always informed by the needs of the Social Sciences and the Humanities.
- As there is no formal or comprehensive definition of research infrastructure for the behavioural and social sciences. The US National Research Council, periodically reconsidering its views on the subject, identifies in its 1998 report two categories:
Category 1: Multidisciplinary centres which provide an opportunity to bring together a critical mass of experts interested in common problems; (a novel variation of the traditional center is the “virtual center” made possible by the internet. )
Category 2: covering a whole range of research instrumentation and equipment, with at least five sub-categories:
2.1. platforms and observational systems,
2.2. computational systems,
2.3. laboratory and analysis systems,
2.4. communication and network systems,
2.5. information systems and databases (e.g. digital libraries, large surveys).
The last sub-category is considered by some to constitute the INTELLECTUAL infrastructure, with a major component being the methodological developments (and computer programs) essential to any sophisticated analysis of the data.
Data Availability
- A detailed inventory of existing data sources was identified as an important aid to comparative researchers to given an essential and better view of who has what data. The draft NESSIE list should therefore be expanded and. The NESSIE stock taking of comparative research should be expanded including both data stocks, and codes of best practice and be given wide visibility. All researchers should cooperate in providing further information, as stocktaking is always difficult even on national basis for all disciplines.
- Identification of future data needs is essential.
- Discussion of data within infrastructures should consider all aspects – data caring, data sharing and data comparing.
Data caring, which also includes data producing, involves constant interaction between data production, concepts, hypotheses, models and theories.
Data sharing has many facets, particularly in an infrastructure perspective. Developments in information and communication technologies and the serious advances being made in documentation and standardization procedures mean that some aspects are already being dealt with. Others require more attention, most particularly the issues of data quality.
Data comparing in the socio-economic fields is not as easy as in the hard sciences. In most cases direct or naive cross-national comparisons cannot be made, and comparability between the data must first be produced. Producing comparability between large and nationally differing data sets can amount to a full-time task and almost a profession. This task is, to a large extent, both undone and unfunded. It is recognised that international databases face additional problems of data comparability arising from differences in national data collection methods and definitions; and that no international body, or union of national bodies, as yet exists to support such an international comparative infrastructure. NESSIE discussions stressed the need that this situation be remedied for European comparative datasets and funding be made available for transnational comparability production.
- There is a particular importance for good quality data for comparative research and those creating data should bear the need for particularly thorough and well constructed documentation of data in mind.
- The Future Evidence Base for Comparative Research: There was broad agreement that the future evidence base for comparative research is much wider than only micro data from academic research and public research, even though survey research will remain an important source. Increasingly registers and process produced data are being used, sometimes combined
- There should be a concerted effort to link socio-economic data needs to EU policy priorities, following the INSPIRE model for spatial data. This has been successfully achieved, individually, for the European Social Survey and the e-living surveys, for example.
- Researchers should be aware of the dangers of restrictive confidentiality regulations that might make access to Europe-wide data even more difficult. They should act as a pressure group to represent the views of academic and private research. Lobbying of politicians and MEPs was suggested as an essential step.
- Lobbying of the national governments and foundations, in our respective countries, to encourage them to invest in the means required for building up, at the national level, the research infrastructures and data bases which will make sense at home and be at the same time building blocks for ERA infrastructures in the socio-economic sciences.
- Restrictive pricing policies can act as a deterrent to open and easy access to comparative data. Researchers should keep a watching brief on this issue.
- Coordination of national, European and international efforts with regard to data access policies is essential.
- Data access arrangements are directly related to data quality improvements; this link should be borne in mind by those involved at both national and international levels.
- A “question bank” of survey questions should be established to allow researchers who require information on a wide variety of topics to locate data easily. This could be based on the model of the CASS Social Survey Question Bank at the UK University of Surrey which has a primarily UK focus at present or the ZA-Qbase, the question database of the Central Archive of Cologne with focus on comparative and German surveys.
- Efforts should be made to ensure that more open access to data produced by the commercial sector is possible. Collaborative ventures between data users and data producers, such as the transnational network of the European Consortium for Communications Research, might serve as a model for such collaborations.
- Access should not only be access to data and corresponding instruments, but also access to research proper and to training for research.
Facilitating Access and Analysis through Technical Development
- Creating comparative data on Europe is both valuable and extremely difficult. Maximum communication between those engaged in the data collection, dissemination and analysis should be encouraged and supported. The efforts of NESSIE in this regard should be continued in the future.
- The descriptions of “best practice” illustrated in NESSIE Roundtables should be widely publicised to ensure that future researchers can benefit from them.
- Technological innovations which support standardisation of survey descriptions should be supported and widely disseminated to allow easier and more efficient resource discovery and more reliable use of those resources.
- The development of standards for data description and metadata should be encouraged and coordinated. The example of the international standard of the Data Documentation Initiative (DDI) and its attempts to remain compatible with the Dublin Core standard was cited as an important model for the future.
- Discussion touched on several important developments for Internet access such as NESSTAR, Madiera, MetaDater. In addition, the implications of developments in E-Science and GRID raised queries on the particular needs of the social sciences which can or could be met by these developments. It was agreed that social scientists are already undertaking activities which are based on the same principles as these developments in several respects.
Harmonisation and Standards
- It was questioned whether, in spite of the existing respectable proposals (e.g. CASMIN, and Goldthorpe), there has been any real progress. The issue of standard classifications should therefore be pursued. The need to maintain national classifications was emphasized, but for comparative research we need common yardsticks over time and over populations.
- Harmonisation activities should always include meta data to improve the inter operability of data bases
- More research needs to be carried out to ensure that data quality and reliability are maintained in all harmonised comparative datasets on Europe – whether input harmonised (ex-ante) or output harmonised (ex-post).
- Full documentation of the harmonisation process must be distributed along with the data. This must include original coding frames, categories, and value labels. In addition, contextual background information is essential for accurate interpretation of analytical results.
- More research should be carried out across harmonisation projects to further investigate the extent to which cross-project standard definitions and variable constructions are feasible and desirable. Specific existing schemes, such as the CASMIN educational classification and the Erikson and Goldthorpe social classification, should be tested on existing datasets to determine their applicability.
- Continuing user support activities for users of these complex data collections are essential, and must be properly funded, so the large harmonised comparative datasets can be exploited to their full potential.
Dealing with Data Protection and Confidentiality Issues
- The new Data Directive provides a useful framework, but social scientists need to carefully monitor the different ways it is being implemented in different states. It should always be balanced with policies assuring access.
- The social science research community should be urged to consider the implications of the European Data Directive as a matter of increasing and urgent concern. There is a need to understand the context in which data exchange will need to take place in the future.
- Consideration of the specific and general needs of the social sciences in terms of data access and data access policies is urgent as the Data Directive is being differentially implemented across the various countries of the EU.
- Consideration of the nature of membership on the Research Ethics Committees under the EU Data Directive, and of the possibility of a harmonised standard for data access and data access policies and perhaps a European standard for minimal requirements for RECs (including international and interdisciplinary representation.)
- Investigation of the implementation and implications of the safe harbour principle for access to sensitive data is essential.
Research Infrastructure
- The definition of research infrastructures for the social science needs urgent and well-founded discussion within the social science community.
- There is a need for coordinated exchange of information and for action between all those involved in data production and analysis. Joint initiatives should be organized and followed through as widely as possible.
- Active involvement of Eastern European countries in all activities needs to be further encouraged.
- The collaboration between the European data archives should be used as a model for the development of the wider research data infrastructure.
- There is an urgent need to establish good links at the national level between the research community and the National Statistical Institutes.
