Abstract
To explain country differences in an analytical or structural dependent variable, the application of a macro–micro-model containing
contextual hypotheses is necessary. Our methodological study examines whether empirical studies apply such a model. We propose
that a theoretical base for country differences is well described in multilevel studies, but aggregate and individual data
analyses fail to specify contextual hypotheses ex ante (in the theory section of an article) and instead elaborate on macro–micro
explanations ex post (in the discussion section). To test our assumptions, we analyzed 22 studies published in journals cited
in the Social Science Citation Index between 2007 and 2010, which compare countries with respect to life satisfaction. Results
are in accordance with our expectations. We conclude that cross-country comparisons should apply a macro–micro-model theoretically
and empirically, if possible, and include the meso level, if appropriate. In case of insufficient data (i. e. only individual
level or aggregate level data), applying a macro–micro-model theoretically may prevent premature conclusions.
contextual hypotheses is necessary. Our methodological study examines whether empirical studies apply such a model. We propose
that a theoretical base for country differences is well described in multilevel studies, but aggregate and individual data
analyses fail to specify contextual hypotheses ex ante (in the theory section of an article) and instead elaborate on macro–micro
explanations ex post (in the discussion section). To test our assumptions, we analyzed 22 studies published in journals cited
in the Social Science Citation Index between 2007 and 2010, which compare countries with respect to life satisfaction. Results
are in accordance with our expectations. We conclude that cross-country comparisons should apply a macro–micro-model theoretically
and empirically, if possible, and include the meso level, if appropriate. In case of insufficient data (i. e. only individual
level or aggregate level data), applying a macro–micro-model theoretically may prevent premature conclusions.
- Content Type Journal Article
- Pages 1-24
- DOI 10.1007/s11205-011-9981-8
- Authors
- Alexandra Nonnenmacher, Institute for Political Science, Leibniz University Hannover, Schneiderberg 50, 30167 Hannover, Germany
- Jürgen Friedrichs, Research Institute for Sociology, University of Cologne, Greinstrasse 2, 50939 Cologne, Germany
- Journal Social Indicators Research
- Online ISSN 1573-0921
- Print ISSN 0303-8300