National Centers for Survey Research Research Paper

1. Social Surveys

Surveys are defined in this research paper as observing societal facts using interviews as the source of information and systematic sampling as the base for selecting information sources (cf. de Heer et al. 1999), and they are quantitative instruments to identify distributions of societal characteristics. Questions or items in a survey are understood as indicators pointing to a latent construct or dimension. Psychological tests, including many marketing studies, differ in two ways from surveys: they focus on characteristics of individuals or small groups but not societal characteristics. In most cases the selection of the respondent is not based on systematic sampling. Survey research is, on the contrary, neither interested in information about a single individual nor in small, identifiable groups. Information gathered on the individual level in a survey will be aggregated across groups and even societies. A population census may share items or question wording with survey research studies. However, census studies are not designed to identify latent constructs. By its name, a census is an investigation involving all members of a given population not a sample.

Survey research, hence, requires among others specific methods and techniques for designing sampling frames, dealing with nonresponse, and constructing and validating measurement instruments which must be in turn suitable for complex statistical modeling across all strata of a society. Survey research is a complex, often high-tech, production process. Consider as an example a National General Social Survey like the American GSS or the German ALLBUS (Davis and Smith 1992, Davis et al. 1994). The target is to collect data from about 3,000 respondents. To do so, one has to contact up to 6,000 households and to identify whether they are eligible and willing to participate. Up to 600 interviewers in 630 sampling points will collect the data. About 50– 100 other staff will back them up, supervise the process, or edit the incoming data. The whole process will be directed and supervised by 5–10 senior social re- searchers. The goal is to collect from each and every respondent all information as accurately as possible. The result will be about 1.2 million different information units (answers to questions, etc. by all respondents).

This complexity of the survey production process requires adequate organizational and professional production structures. Big survey research companies like Gallup, Harris, Infratest-Burke, Inra, or Mori to name but a few and famous, provide such organizational and professional structures for polls, big market surveys, etc. In the public, nonprofit domain, independent, academic survey research centers serve the academic communities, governments, and even compete with private companies on many sectors.

It is the integration of academic methodological standards and perspectives into the applied survey research process, which makes academic survey research centers different from private companies’ research units. Publishing rigorous methodological re- search, going beyond the ‘it works’ to find out ‘how and why it works,’ is thus a must for an academic survey research institute. Hence it is no coincidence that the two most prominent survey research centers today, ISR and NORC, are also centers of methodological excellence. However, methodological research, important as it may be for survey research, is a very small area in the sea of modern social research. It is almost always interdisciplinary and cooperative. In short, methodological survey research is not suited for a career at a university. National research centers by their very size and their mission to accomplish high quality surveys provide a natural base for long-term high quality methodological research also.

National centers for survey research cover national and international issues, which are quite distinct from regional centers. In addition such centers have a broad focus on methods, target audience, and topics. Funding and organization differ from center to center and from nation to nation.

The term ‘national center’ as it is used here does not exclude several centers within one nation, nor does it imply that the word ‘national’ in the name immediately qualifies an institute to be a national center. They also serve as hosts for important national surveys. Studies like the General Social Survey (NORC), the Panel Study of Income Dynamics (ISR), British Election Studies (NATCEN), or the German General Social Survey (ZUMA /GESIS) are cornerstones of their countries’ social science information systems. Finally, they have an enormous impact on the development of multinational survey research: NATCEN, NORC, ZUMA are founding members of the International Social Survey Programme (ISSP, www.issp.org), ISR hosts the World Value Survey.

Rigorous methods, the public as an audience, free distribution of major nationwide data sets, intensive, published methodological research, and an international perspective are what mark national centers for survey research.

Today, 88 years after the first survey (de Heer et al. 1999, p. 28) centers which truly qualify are still very small in number. This is despite the enormous growth in interviews conducted around the globe (in 1978 the gross revenue of the survey industry was estimated at $4 billion; de Heer et al. 1999, p. 32); the annual turnover reported by ESOMAR in 1998 was about $13 billion with an increase of 10 percent since 1997 (www.esomar.nl).

There are actually only a handful institutions worldwide which can be counted as a national center for survey research: in the US the National Opinion Research Center (NORC, Chicago, founded 1941), the Survey Research Center in the Institute for Social Research (ISR, Ann Arbor, 1946), in Great Britain the National Center for Survey Research (NATCEN (formerly SCPR), London, 1969), the Norwegian Social Science Data Services (NSD, Bergen, 1971), the German Social Science Infrastructure Institutes (GESIS, Mannheim, an umbrella organization for ZUMA in Mannheim, Zentralarchiv in Cologne and the Social Science Information Center in Bonn, 1986).

Other institutes like social science data archives, for instance the Roper Center, ICPSR, or the Essex Archive (cf. CESSDA, www.nsd.uib.no/cessda/ or www.ifdo.org), or the numerous social research institutes listed in clearing houses (www.abacon.com/sociology/soclinks with references to other lists) either do not carry or contract out major national surveys or are not heavily involved in methodological survey research apart from secondary data analysis. Hence they are not covered in this section.

In the following, short bios of the four centers will be presented. These bios will not go into the history of these institutes; rather their current standing will be emphasized. In concluding, the impact to these centers on survey research will be outlined at the beginning of the twenty-first century.

2. NORC: National Opinion Research Center, Chicago

NORC was established in 1941 as a nonprofit corporation through a grant from the University of Denver and the Field Foundation. In 1947, NORC moved to the University of Chicago campus (www.norc.uchicago.edu). The center conducts survey research in the public domain as well as for private organizations. It is an independent nonprofit organization with strong ties to the University of Chicago.

NORC has a reputation for both complex surveys and innovations in survey research methodology. It is the birthplace of surveys conducted for the benefit of the sociological profession at large, the General Social Survey (GSS) (Davis and Smith 1992). The first GSS was conducted as early as 1972. A GSS is a generalized tool which repeatedly measures important characteristics. It is released immediately after completion to the social science community. Thus researchers, students, and other interested parties have up-to-date survey data to hand which they can analyze on their own terms. Such a complex, long-term enterprise requires a strong and well-equipped infrastructure, like the one provided by NORC as a national survey research center. Other strongholds of NORC are methodological studies concentrating on measurement error. Norman Bradburn and Roger Tourangeau are but two among the number of researchers who contributed to this field from the 1980s.

3. ISR: Institute For Social Research, Ann Arbor

Shortly after NORC moved to Chicago, in 1948, researchers from the Survey Research Center and the Research Center for Group Dynamics founded ISR in Ann Arbor (www.isr.umich.edu). From its very beginning ISR advocated interdisciplinary research. Its continuous studies range from National Election Studies to the Panel Study of Income Dynamics. It provides major databases as well as substantial contributions in group dynamics and political science to the scientific community. ISR is also a world leader in survey research methodology and statistics. Researchers like Kish, Schuman, Converse, Presser, Groves, Couper, Singer, and Schwarz pave the way for modern survey research processes in areas such as sampling statistics, response error, householdnon-response, and survey cognition processes.

Another impact area of ISR is training and data publishing. Since 1962 ISR has hosted the Interuniversity Consortium for Political and Social Research (ICPSR) which was founded as an archive for quantitative social science data. From a safe haven for key punched cards ICPSR developed the world’s largest survey data publisher distributor where researchers have access to thousands of data files.

ISR offers three training programs open to researchers worldwide. These are the Summer Institute in Survey Research Techniques, the ICPSR Summer Program in Quantitative Methods and, more recently, the Research Center on Group Dynamics Program in Experimental Methodology and Statistics. The institute puts much emphasis on international cooperation and knowledge dissemination in all its activities.

Apart from a healthy competitive attitude, NORC and ISR are complementary institutes with strong ties to their respective universities (NORC—University of Chicago, ISR—University of Michigan at Ann Arbor). NORC puts more emphasis on fielding many complex studies, while ISR concentrates on specific studies. ISR is very strong in survey methodology, while NORC is stronger in survey process management and sampling design. Both provide a base for international cooperation.

4. NATCEN: National Centre For Social Research, London

NATCEN, formerly SCPR, was founded in 1969. Today, it is Britain’s largest nonprofit survey research organization (www.natcen.ac.uk). It is the home of British Social Attitudes as Britain’s general social survey, of the British Elections studies, and major health surveys, to name but a few. Within the tradition of British social research, NATCEN conducts both quantitative and qualitative studies. Among its divisions are a Survey Methods Centre (research, consulting, and teaching) and joint centers with universities like CASS (Centre for Applied Social Surveys) which hosts a survey item bank. Researchers at NATCEN are leading in applied survey research. Among their innovations is the International Social Survey Programme (ISSP) initiated by Roger Jowell and the first ever deliberative poll. It will be the ‘command and engine room’ of the European Social Survey to come in 2002.

NATCEN is the only center among the survey research centers discussed here which names itself a ‘national center.’ However, it is a national center not by Her Majesty’s appointment, but de facto due to the specific structure of the British science funding system. As an independent charitable trust it forms the base for collaborative mid-range and long-range projects like CREST, the Centre for Research into Elections & Social Trends. Funded as a research center of the Economic and Social Research Council (ESRC), it is a joint venture of Nuffield College in Oxford and NATCEN in London. The above-mentioned CASS, on the other hand, is an ESRC research center joining NATCEN with the University of Southampton.

5. GESIS: German Society Of Social Science Infrastructure Institutes, Mannheim, Cologne, And Bonn

Founded in 1986 by act of the federal government and all states, GESIS comprises three major centers: ZUMA, Center for Survey Research and Methodology in Mannheim, the Zentralarchiv in Cologne, and the Information Center (www.gesis.org). Together these three centers form a comprehensive national survey research center as a permanent infrastructure for the social sciences.

ZUMA consults academics in all areas of quantitative survey methods; it runs continuous research streams in areas like standardization of demographics, intercultural survey research, and, recently in web surveys. It is, among others, the home of ALLBUS, the German General Social Survey, the German ISSP, and a participant in the German Social Welfare Studies. Together with researchers from five other European institutes it will be part of the European Social Survey ‘engine room’ (see above). It is also the birthplace of the German Standard Demography, now a major project of the German Census Office, ZUMA, and other professional groups. The center bridges the gap between German Official Statistics and the social science profession by providing and/or enabling access to official microdata.

The Zentralarchiv, founded as early as 1960, is the German national archive for quantitative survey data. It played a major role in establishing the international archive network. The archive also provides training courses in survey methods.

The Information Center provides in-depth bibliographies of German social science publications, keeps record of ongoing research projects, and serves as a platform for information interchange between Eastern and Western Europe.

Taken together the three centers form a unique infrastructure for their national scientific community. Access to all services is free, products are nominally charged.

The three institutes were independent and differently funded units until the National Science Board (Wissenschaftsrat) voted to form a unified national survey research center, namely GESIS. As a social science infrastructure it has a strict supportive function for academic research. Substantive research is different than in America or Britain, and is not a core issue of GESIS. However, as a dedicated social science infrastructure GESIS is fully funded by the federal government and all German states.

6. Conclusion

National centers for survey research are defined as institutions that provide databases, methodological research, and training for their scientific community. In addition, national centers focus on the national rather than the regional level and bridge national research with international. There may be more institutions around the globe than the four introduced here. Many others may fulfill one or other of the functions. However, the core of a national center for survey research is dedicated methodological studies embedded into long-term methodological and substantial research programs.

In looking back to get a vision of the future, one could arguably say that the fast progress one could observe in survey research technology and methodology since the 1950s was caused by the very existence of the two American centers. Others countries still lag behind the United States both in advancement of scientific knowledge and in survey engineering. Germany, for instance, where ZUMA was founded as late as 1974 with limited capacities (no fieldwork staff, no own surveys), had to bridge a huge methodological and technological gap since the 1980s. The founding of GESIS as a national center was a late but not too late remedy. It is thus no coincidence that the lack of national centers or their substitutes correlates with the standards of survey research in a country. Because all modern societies need sound information about their own state, efforts should be made to either found more national centers or to transform existing ones into real or virtual transnational centers. Finally, in view of the new survey technologies, like net-based multimedia surveys, and their likely abuse and misuse, national centers become even more indispensable as a means to sort good methods and data from information rubbish.

Bibliography:

1. Davis J A, Mohler P P, Smith T W 1994 Nationwide general social surveys. In: Borg I, Mohler P P (eds.) Trends and Perspectives in Empirical Social Research. Walter de Gruyter, Berlin
2. Davis J A, Smith T W 1992 The NORC General Social Survey— A User’s Guide. Sage, Newbury Park, CA
3. De Heer W, De Leeuw E, van der Zouwen J 1999 Methodological issues in survey research: A historical review. Bulletin de Methodologie Sociologique 64: 25–48