Katharina Meitinger (GESIS - Leibniz Institute for the Social Sciences, Mannheim, Germany)
Eldad Davidov (University of Cologne, Germany, and University of Zurich, Switzerland)
Peter Schmidt (University of Giessen, Germany and Cardinal Wysczinski University Warsaw, Poland)
Michael Braun (GESIS - Leibniz Institute for the Social Sciences)
Articles due by August 31, 2018
There has been a tremendous increase in cross-national and longitudinal data production in social science research in recent decades. Before drawing substantive conclusions based on cross-national and longitudinal survey data, researchers need to verify whether the measures are indeed comparable.
Measurement invariance tests are an increasingly popular way of assessing the cross-national comparability of survey data. However, researchers often find it particularly difficult to establish the highest level of measurement invariance, that is scalar invariance. As a consequence, latent means of constructs and composite scores cannot be meaningfully compared across countries and several analysis strategies tailored to analyze comparative survey data, such as multilevel or multi-group analyses, cannot be performed with confidence.
In recent years, the predominant approach to "fixing" this issue has been to opt for more statistical sophistication and to relax certain requirements when testing for measurement invariance. Approaches like approximate measurement invariance by Bayesian structural equation modelling (BSEM) or alignment fall in this category. These approaches are promising because they are more liberal. As such, they may often suggest that measurement invariance is given while more traditional, stricter approaches indicate that it is absent. This means that they may allow researchers to perform meaningful comparative analyses more frequently.
However, these approaches cannot provide reasons for the absence of measurement invariance. An alternative approach in this context is to view the lack of measurement invariance as a source of information on cross-group differences and to try to explain the individual, societal or historical sources of measurement nonequivalence. On the one hand, quantitative approaches-such as the multiple indicators multiple causes model (MIMIC) and multilevel structural equation models (MLSEMs)-aim to substantively explain sources of measurement non-invariance. On the other hand, there is an increasing awareness of the potential of mixed-methods approaches to explain instances of measurement non-invariance by combining measurement invariance tests with different qualitative or quantitative approaches.
We hope that the special issue will provide a forum for discussion on these innovative approaches to survey data and will help researchers in their endeavors to conduct meaningful comparative research.
For the special issue in Survey Research Methods we invite studies that:
To submit, go to the SRM website and upload the article as a PDF, just like a standard SRM article; be sure to mention the special issue in the field for Author comments
Reviewing policies losely follow the usual standards of SRM: Each submitted paper will be assigned to one of the three editors of the special issue. The supervising editor will then select two expert reviewers. Those reviews and the reading of the supervising editor will be used for the final decision.
Researchers who want to submit are requested to adhere to SRM's implementation of the Transparency and Openness Promotion guidelines of the Center for Open Science. See the Author Guidelines for a detailed description of those policies.