Mikael Hjerm, Sociology, Umeå University

European Social Survey

2009-2012

ESS is an attitude and behavioral survey that has been undertaken five times in more than 30 European countries, including Sweden. Two to three thousand individuals per country are interviewed about their attitudes and behavior towards, for example, immigration, democracy, equality, media and medicine use, participation in the civil society, trust, health, networks and psychological predispositions. The data collection in the fifth round of ESS ran during 15th of September 2010 through 15 of March 2011. The importance of understanding people's attitudes, values, and behavior has become increasingly important in order to be able to understand contemporary society in a time of political and societal change. ESS has in 10 years time been established as the most important and reliable source of information concerning attitudes, values and behavior for scientists, mass media, political decision makers and analysts. Factors like the spectrum of the examined topics, the solid guarantee of quality, standardization, documentation, the swift production of data and, not the least, the general access to data has contributed to the success of ESS.

ESS has three overarching scientific goals. First, to enable comparison of key indicators within and across countries combined with the examination of current attitudes and preferences. The aim is to produce the objectively best survey data combined with a never before seen content multiplicity. A proof of success is that ESS has, as the only social science project ever, been awarded the prestigious Descartes prize for scientific excellence.

Second, to be able to measure changes in social phenomena. The majority of the questions are therefore repeated in every round of ESS. Many changes in attitudes and behavior take a long time, which means that we have not yet witnessed the full significance of ESS for our understanding of social change.

The third goal is to create sustainable social indicators that can be used to describe the state of things across time and countries. Indicators that deal with, for example, equality, crime, and happiness are constructed to work complimentary with established economic measures like GDP. These indicators can be used by researchers, politicians, and analysts to compare countries and regions or groups of individuals in an established way.

The cost of the project is primarily related to the data collection. RJ's part of this equals 1/3 of the total cost. The two other parts were financed by FAS and VR. The data collection was planned and undertaken during spring 2010. The planning included contraction the fieldwork organization (Stat Sweden), participation in the general development of the survey, translations of the survey and planning data collection more precisely. The survey was tested during August 2010 and the fieldwork was undertaken between 15th of September 2010 through the 15th of March 2011. Stat Sweden made the final delivery of the data file in the end of March the same year. The data was quality checked and tested between March and May before delivery to NSD in Norway. The data was further tested by NSD consulting the Swedish team between May and august before publication in October 2011. After publishing the final data file for ESS5 the planning of ESS6 started.

The Swedish part of ESS5 fulfilled all the criteria put up by ESS as of technical demands. The response rate of 51% was considerably lower than expected. We are still not sure about the reason for the latter as we are still examining those reasons. We used all strategies know to enhance response rates, like reissuing refusal to the best interviewers, using incentives, new modes of contact etc, but those measure had very limited effect. Due to the general trend of declining response rates we have initiated mixed mode experiments that will take place during round 6 in order to examine if that can enhance response rates as well as having positive effects on the total cost of the project.


The survey ESS5 contained the following parts
Communication [10 questions] The extent and type of mass media people use as well as items on trust
Politics [app. 40 questions] Party preferences, voting, political participation, political values as well as welfare state attitudes, xenophobia and the role of government.
Wellbeing and belonging [app. 36 items] Ethnic and religious belonging, religiosity, social contacts and networks, happiness, health, security, crime and terrorism
Sociodemographic profile [app. 73 items]
Household composition, sex, age, education, income, work. Employment relations, working life for the respondents and also much information on spouses and parents.
The Schwartz scale [ 21 items]
The Schwartz scale is an established scale to classify respondents in accordance with their fundamental value orientations.
Test items [app. 30 items]
Replications of items in the same survey, similar items as previously asked but with differences in response scales and formulations.
Interview questions [app. 10 items] The interviewer's assessment of the interview and her apprehension of the respondent's willingness to comply with her best knowledge.
The rotating modules in round 5 covered
Work, Family and Well-being: The Implications of Economic Recession

The Implications of Economic Recession. This module will repeat many of the items from the 'Work, Family, and Well-being' module included in ESS2.
Trust in Criminal Justice: A Comparative European Analysis This module will deal with the European's view of criminal justice.

The national surveys are translated from English. Every country makes at least two independent translations that are commented upon by a third person and then judicated by a fourth person. In Sweden we only interview in Swedish, but in countries were any given minority exceeds five percent of the population interviewing is also done in the minority language. The process of translation takes place in dialogue with the expert translation groups within the ESS that make instructions for specific translation issues as well as handles cross national problems. In round 5 the translations were also evaluated by an external team as well as by special software (cApStAn) in order to further strengthen the quality of the translations. This process guarantees maximum comparability across countries.
Sampling is done by random sampling of the total population above 15 years of age in Sweden and countries where random individual sampling is possible. Household registers are used in countries were individual sampling is not possible. No substitutions or similar tampering is allowed. The gross sample is approximately 3000 individuals in all countries. All sampling procedures and sampling frames are specified in detail and monitored by the ESS central coordination team.
Data collection The mode is face to face interviews in all countries. The data collection is preceded by national test surveys in order to secure translations, questionnaires, and programming. Things like the amount and timing of contact attempts and strategies for refusal conversion are examples of standardized and documented issues within the ESS.
Post handling of data is done in each country in accordance with standardized modules. The latter includes, for example, the coding of occupations according to ISCO88 and standardizing levels of education in accordance with the ESS standard. The post coding in Sweden is done by the national team in cooperation with SCB. The finalized data files are then quality controlled before delivery to Norsk Samfunnsvitenskapelig Datatjeneste (NSD). NSD makes the final quality control with specially designed software combined with manual checking before data is made public.
Publication of data. Data is made public to everyone at the same time at www.europeansocialsurvey.org

Usage
ESS is today agenda setting when it comes to international surveys on account of the quality. ESS is today the number one organization that provides international comparative survey data to researchers around the world

ESS already has more than 40,000 registered data users. A key goal of the ESS is to use its authority and expertise to help advance methods of quantitative social measurement and analysis in Europe and beyond. It strives to do this not only by example, but also by an innovative programme of methodological research and seminars designed to inform academic and professional practice, and by an ambitious training programme in research and analysis techniques, which has been heavily over-subscribed from the start. The ESS website facilitates access to the datasets, to the online data analysis tool (including a multi-level data repository and an e-learning programme). The ESS Bibliography is an online facility through which researchers may review the different publications; where possible, these are linked to the Bibliography so that it functions as a repository for data users and analysts.

Conclusion

ESS is a pure infrastructure project, which means that the scientific goal for the Swedish participation is twofold. First, to collect high quality data and, second, to actively contribute to the development of ESS. Both these goals have been fulfilled. The first goal via the successful data collection and the second goal via active participation and involvement in ESS. The latter includes instigating changes of specific question as well as contributing to between ways to register contact attempts with the respondents and so on.
We have by participation in the ESS provided data on the Swedish population that can be analyzed separately or together with data from other European countries, by scientists, political decision maker, journalist, or an interested public. Given the enormous downloads of the ESS dataset and the massive body of publication it is obvious that ESS is a very successful infrastructure project. The Swedish participation in ESS is just one part, but ESS stands and falls with active participation from the European countries. We have fulfilled our part within this framework.