On advantage of using social indicator data, such as the National Crime Survey (NCS) or the Uniform Crime Reports (UCR), is that these large-scale data sets allow researchers to identify the correlates of offending. The NCS data can identify the high-rate offending subgroups among the available demographic characteristics. There is substantial agreement across official and self-report data regarding the age and six of offenders. To facilitate etiological research that focuses on the intervening variables between demographic correlates and delinquency/crime, research should merge analyses of social indicator data (UCR, NCS) with special indepth studies of individual subgroups. One method that could be adapted for this purpose is oral histories of offenders and victims (Laub, 1984). Oral histories are particularly helpful in identifying new variables for analysis, and they also enhance the study of the interaction of key variables such as age, race, sex, and social class. Oral histories permit a study of the social process, i.e., the ordering of events over time, the duration of events, the intensity of events, etc. 3 tables and 2 figures.