Seminar “Quality and Comparability of Measures for Constructs in Comparative Research: Methods and Applications”

Začetek: Junij 10, 2009
Konec: Junij 13, 2009
Kraj: University of Bolzano/Bozen (Italy)
Organizator: Quantitative Methods in the Social Sciences
This seminar deals with the problem of the comparability of subjective and objective measures for theoretical concepts in cross–national and longitudinal research. In comparative research means of variables and relationships between variables are compared across countries. The question is if these comparisons can be made: What guarantees that the measures used are comparable across countries and over time?

In the methodological literature criteria such as “metric” and “scalar invariance” have been formulated for these comparisons. These criteria are very strict and this may mean that comparison is only possible on a very limited scale. So the first question is to what extent do comparisons become impossible if one uses these criteria? Furthermore, should one be so strict? “Partial invariance” was also proposed in the methodological literature, but it is not clear if it is sufficient for comparative research. Furthermore, it is also interesting to see how far substantive researchers take these criteria into account or how they evaluate the comparability of their measures for these comparisons. Examples for situations where researchers should consider these criteria are mean comparisons of measurements across cultures, comparison of regression coefficients across countries or time points and applications of multilevel models.

In the research of background variables, such as income, education and occupation, special attention has been paid to the comparability of the measurement of these variables. In this context “harmonization” of the coding procedures plays an important role. However, one can still raise the question whether “harmonization” guarantees comparability across countries.

Another problem that plays a role in the comparison is the quality of the measures for the concepts in the different countries. If in one country the measurement errors are larger than in another then the results across countries are also not comparable.
In this seminar, we want to bring together methodologists and substantive researchers working in different fields to get a good picture of the used (in practice) and required procedures for evaluating the comparability of measures across countries.

The idea is to have four sessions of half a day. In these sessions the applied procedures for different topics will be discussed by methodologists in that field and substantive researchers.

The topics we suggest for the substantive sessions are:
1. Schwartz Human Value scale (Organiser: Eldad Davidov)
2. Social and political Trust (Organiser: Patrick Sturgis)
3. Media use or exposure to specific topics in the media (Organiser: Willem Saris)
4. Background variables: occupation, income and education (Organiser: Harry Ganzeboom)

In each session three papers will be discussed from specialists in this field. These papers are expected to be sent around before the meeting. The discussion will take place between the presenters in the session and the other participants in the seminar. In this way there will be also the possibilities that the specialists in the different fields learn from each other and from researchers in other fields.

Other participants

Applications are now invited for attendance at this meeting. Participants will be selected on the basis of their experience with one of these topics or their methodological expertise. They will not have the opportunity to present a paper but will get all the papers and have the opportunity to discuss these issues with the other participants in the different sessions.

Applications for attendance are to be sent to the email address of Eldad Davidov (davidov@wiso.uni-koeln.de) no later than April 15. Seminar applicants should submit half a page (not more) outlining their experience and why they wish to attend, and a list of publications together with their application. Applicants should state if they have previously attended a QMSS summer school or a QMSS seminar and how they will benefit from participation in their future research as the seminars are intended to provide a vehicle for continuing collaboration and networking.

Other information
There is no registration fee, and accommodation and meal costs during the seminar will be covered. Participants need to pay for their own travel, but the organizers will contribute to the travel costs of up to10 participants.

Organizers: Willem Saris (w.saris@telefonica.net), Universitat Ramon Llull
Eldad Davidov (davidov@wiso.uni-koeln.de), University of Cologne
Sylvia Kritzinger (sylvia.kritzinger@univie.ac.at), University of Vienna