There are several ways to measure crime, and several sources of crime and victimization data. Each has its own strengths and weaknesses. More importantly, for the question below, the specific purpose of the research will determine what types of measurement and/or sources of data are appropriate. As an example, if I wanted to study homicide, a victim survey would be inappropriate because the victims are deceased. For this particular research purpose, the Supplemental Homicide Reports (SHR) would be my best source of data.
Think about the city where you're from. If you wanted to evaluate a community policing program that encourages residents to report incidents to the police, how would you measure crime in that specific city?