How do we analyse crime statistics? Lucy Snowball NSW Bureau of Crime Statistics and Research
About me Bachelor of Arts/Commerce –Statistics, maths and economics Statistician at the Bureau of Crime Statistics and Research –Use statistical methods to analyse crime and offending patterns
About BOCSAR Data from police and court records Types of research –Who gets sentenced to prison? –Juvenile offending –Social and economic factors that affect involvement in crime –Predict re-offending
Methods we use Summary statistics –Mean, standard deviation, median, mode, range Time series analysis –Increasing/decreasing trends –Break points Regression modelling –Building an equation to see which factors influence others
Factors to consider How is the offence recorded? Which groups are most likely to commit this offence (young people, men, people in cities) What effect will policing policies have? What external factors might affect rates –The economy, government policies, commercial decisions, other crime rates
Example: Narcotics use Heroin is the most common narcotic Police record each time they arrest a person for use or possession of a narcotic Offenders: young people (18-30), people in major cities (Sydney, Newcastle, Wollongong) The time series can reflect actual use or policing strategies All drug use is affected by the availability of that drug Effects other crimes - Narcotics linked to property crime