The Perils of Periodical Indexes

Abstract

This article is about a topic neglected in the social science literature on content analysis and culture indicators methodology—problems in using periodical indexes to construct research samples. The author briefly identifies and discusses reasons why article headlines located through indexes' subject categories can prove to be misleading indicators and then turn to the methodological difficulties arising from historical and idiosyncratic changes in index subject category headings and subheadings—difficulties particularly relevant to longitudinal research. The author argues that researchers using an index's subject categories should test not only for category longevity but also for category coherence and consistency over time. The use of oversampling, cross-references, and other devices is suggested as a means to correct or compensate for hidden inaccuracies in index classification and to construct purposive samples for analytic comparisons.