In professional or real life research it is hard to conduct a probability sample or even a random sampling of a specific population. However, as a researcher we may have access to all offenders on a sex offender registry in a given state, or a list of juveniles in a specific treatment program. Let me know if this makes sense