Repeatable and will provide a similar result if done again. A structured scientific method will be able to be repeated. Identifies trends, patterns and comparisons.
How possible it is to generalise the results of the findings to the general population. It depends on the sample (who you are researching). A larger sample is more likely to be representative, but does it reflect the main population?
Official statistics allow criminologists to see exactly what crimes are reported to the police, as well as certain details, such as the gender of a criminal, and their ethnicity
Self-reporting studies have issues as a perpetrator may not want to reveal the crimes they have committed themselves, or make them seem to commit more crimes than they actually have