| Quantitative analyses of crime data are often hindered by data limitations. In order to maximize sample size to utilize certain analytical techniques, researchers have often used cross-sectional research
Quantitative analyses of crime data are often hindred by data limitations. In order to maximize sample size to utilize certain analytical techniques, researchers have often used cross-sectional research designs based on convenience samples of all countries that participate in a particular data collection. This approach has been criticized for two main reasons: (1) cross-sectional designs are often used to test theories of temporal variation, and (2) sample size is maximized at the expense of theoretical relevance and socio-historical context. One response to these critiques is a longitudinal analysis of a smaller, theoretically meaningful sample that allows for comparisons both across time and countries. This paper discusses the methodological issues involved in analyzing such a sample, including compiling crime data from a variety of official records sources, operationalizing independent variables, integrating qualitative and quantitative data, adopting an interdisciplinary approach to account for socio-historical context, and choosing appropriate quantitative methods to analyze time-series-cross-sectional data for a relatively small sample. This paper illustrates the benfits and limitations of analyzing a small sample of countries selected purposively rather than conveniently. Examples are provided from a longitudinal study of social change and crime in Central Eastern Europe. |
Updated 05/20/2006