2 ms·

Per capita would hide lurking variables like differences in crime rate among different communities. Comparing to overall homicides surfaces this effect more.

• thebooktocome 10y ago
  That's just not how you correct for confounding effects.