At the start of the study, there were no statistically significant differences
between the Triple P and control
group counties in observable characteristics (e.g., child maltreatment rates over the
prior five years, poverty rates,
racial composition).