The research was conducted between 1989 and 1991 in three probation services. It had six main objectives, including determining whether the report content, sentencing decisions, and probation officer recommendations varied according to ethnicity and gender considerations. The research initially encountered a lack of data, especially on women, and problem in persuading probation personnel to take part by completing questionnaires on offenders' ethnicity. In addition, the three probation services used three different classification systems. The experience revealed that although, ideally, recorded ethnic categories should correspond to the self-perception of the offender or victim and suit the purposes of the research, in practice the researcher may have no choice other than to use the classification system of the agency being researched. In such cases, it may be useful to consider the classification system itself, because it may reveal much about the form of racism at any one time. The experience also revealed issues related to being a white researcher focusing on racism. Notes