ABSTRACT
Research on political orientations and behaviour confronts similar problems. T h e t e rm 'Burgerinitiative' is closely related to, but certainly not identical with, the Dutch phrase ' inspraak' , and comparisons of levels of par t ic ipat ion should take these differences into account (van De th 1986). Similarly, the measurement of concerns for pollution and ecological issues should not be based on the same set of items or topics in different countries since national restrictions and subsidies vary regarding car catalysts, wrapping, drainage water, and garbage discharge (Nas 1995). A crucial p rob lem of comparat ive research, then, is the establishment of cross-culturally or cross-nationally valid and reliable instruments. If, on the one hand , we use culture-or nation-specific indicators for our concepts, we have problems tracing cross-cultural or cross-national differences. If, on the other hand , we construct identical instruments for various settings, we are unlikely to obtain an appropr ia te measure of nat ional or cultural phenomena . How, then, are we to sail between the Scylla of losing national or cultural validity and the Charybdis of endanger ing cross-cultural or cross-national comparabi l i ty? 2
T h e p rob lem of finding different indicators to compare the same concepts or p h e n o m e n a is obvious to every novice in the field of comparat ive research, and m a n y textbooks or overviews underl ine the impor tance of using equivalent instead of identical indicators . 3 Yet despite the striking consensus about the impor tance of the problem, the n u m b e r of proposals to deal with it are ra ther limited, and a t tempts to handle equivalent indicators in comparat ive research are the exception, not the rule. This situation is due first of all to the complicated logical and empirical aspects of the identity-equivalence problem: that is, 'establishing credible equivalence is difficult, as "mean ing" is contextual ' (Teune 1990: 54). Indeed, the 'problems are severe and it is easier to explicate them than to suggests ways of dealing with them ' (Verba 1969: 64).
