Public Opinion about the Environment: Testing Measurement Equivalence across Countries

Abstract

Environmental issues and opinions are increasingly recognized as topics of importance to a global audience. Scholars in the social sciences seek to describe trends in public opinion about environmental concerns cross-nationally, yet use a variety of measures and analytical techniques for such work. Although an understudied topic, the question of measurement comparability across contexts is important for this line of inquiry. This research examines public opinion on the environment in a cross-national context using the 2010 International Social Survey Program (ISSP) Environment data. I use structural equation modeling with latent variables to construct a latent measure of environmental risk perception across 32 countries. I examine the composition, level, and distribution of the latent construct environmental risk perception cross-nationally, with comparisons across regional groups. Particular attention is given to how structural equation modeling can be used as a tool for testing measurement equivalence in cross-national research.

Keywords:

• cross-national
• environmental attitudes
• ISSP
• risk perception

ACKNOWLEDGMENTS

The guest editor of this issue, Markus Hadler, handled the review process for this article. The author would like to thank the anonymous reviewers for their helpful comments on this submission.

Notes

This classification scheme is an attempt to bridge the expanding sample of countries included in the ISSP data with both the previous versions of the ISSP Environment Data from 1993 and 2000 and prior work on environmental concern’s steady use of the division of countries into developed/industrialized and developing/industrializing countries as groups. As such, for the 2010 data, a few placements, such as South Korea and Taiwan in the developing countries group, stand out.

Important assumptions of the χ² test include (1) the absence of outliers, (2) that the multivariate distribution of the observed variables is normal with respect to kurtosis (Browne 1984), (3) the sample is sufficiently large for the asymptotic properties of the test, and (4) the sample comes from a single population. (For more detail, see Bollen 1989; Paxton, Hipp, and Marquart-Pyatt 2011.).

More detail information is available upon request from the author. There is a group of countries within each region for which a slight difference can be discerned. The question of the cross-group relevance of this is yet to be examined. A cluster analysis would be a fruitful extension.

Additional information

Notes on contributors

Sandra T. Marquart-Pyatt

Sandra T. Marquart-Pyatt is an associate professor in the Department of Sociology and Environmental Science and Policy Program at Michigan State University and editor of the International Journal of Sociology. Her core areas of research interest include cross-national research on environmental issues, outlooks and concerns, sustainability, and democracy.