Do We Trust the Data? On the Validity and Reliability of Cross-National Environmental Surveys
Abstract:Objective. This article provides evidence on the validity and reliability of cross-national environmental surveys, an aspect that has not found much attention so far. Methods. Validity can be checked in examining whether cross-national differences in environmental attitudes are in accordance with theoretical hypotheses. For example, a study can claim some validity if concern about inadequate sanitation is strongly negatively correlated with the actual extent of access to sanitation in a country. Results. Several validity checks were undertaken for the Gallup, Gallup, and Dunlap (1993) survey, all of which tend to support its validity. Next, the reliability of cross-national environmental surveys was checked. Both Pearson and Spearman rank correlations were run for similar questions from differing studies. Most correlations were low and statistically insignificant, however, thus putting some doubt on the reliability of cross-national environmental surveys, at least with respect to the questions examined. Conclusions. The findings support the validity of cross-national environmental surveys, but not their reliability. Future surveys should be designed such that validity and reliability checks become easier to undertake.
Document Type: Research Article
Affiliations: London School of Economics
Publication date: March 1, 2002