Tuesday, August 14, 2012

What Are Accurate Crime Stats Worth To You?

The South Carolina news outlet The State had a story about Columbia Police spending nearly a quarter of a million dollars to upgrade their crime records system. There was this great bit on the reasoning behind the expenditure:
Accurate crime data is essential to public safety. It allows police to know where to focus resources. It helps City Council decide what ordinances and programs are needed to improve public safety. And reliable data provides a benchmark that lets the public know whether those efforts have been effective.
 I couldn't agree more. What is your agency doing to ensure that your crime stats are accurate?

Read more here: http://www.thestate.com/2012/08/06/2386281/columbia-police-taking-steps-to.html#.UCQckqP-2kI#storylink=cpy

1 comment:

  1. Hi Scott, we're following the CompStat model, so numbers mean everything. However what I've noticed is that numbers are only skimming the surface. Because I was in Planning and Research, I've been pegged as the "stats" and "computer" guy. That's the problem though is that it means I haven't done analysis. Because of my role in Planning (I've now changed over to Patrol), I was responsible for just producing the statistics. Now as an analyst for patrol, I can't simply see numbers and know if crime is up or down. As an analyst, my view is to put those numbers into context and interpret what they mean. Then I want to move from there and change the information into knowledge (i.e., actionable).

    The reason why I asked you in one of your previous posts about reported vs. occurred date is that we're currently reporting on incidents based on reported date. What's happened a few times is that our stats looked skewed for this reporting period because of a series/spree that occurred last period, but reported in the current period. It happens especially for property crime. When I'm performing an analysis to see if there are patterns, I base my analysis on when the incident occurred rather than when it was reported. I believe this is more accurate. Also if executive staff are trying to figure out if we've got a problem, I recommend using when the incident occurred. Currently our stats are based on when they were reported. I'm not sure if this is accurate. What are your thoughts?

    John

    ReplyDelete

I reserve the right to remove defamatory, libelous, inappropriate or otherwise stupid comments. If you are a spammer or are link baiting in the comments, a pox be upon you. The same goes for people trying to sell stuff. Your comment will be deleted without mercy.