In no more than a decade, the multilevel model has spread throughout social sciences, becoming one of the basic techniques in the discipline. The focus of this article lies on the use of multilevel analysis in the domain of international comparative studies. The article shows a broad picture of potential pitfalls in using multilevel analysis for the comparative researcher. The increasing popularity of the technique does entail some potential dangers researchers are not always aware of. Five of them are dealt with in the article: definitions of multilevel comparative research, the correct use of units and levels in comparative analyses, the comparability of instruments, the availability of data and indicators and stepwise multilevel analysis strategies. The conclusions places the potential pitfalls back in a broad perspective and sketches potential directions and promises of multilevel comparative research.
Keywords: Multilevel analysis, secondary data analysis, comparative research, quantitative research, sociology of health and illness, analytical model.