Friday, July 20, 2012

60 Steps Revisited: Step 26 - Take Account of Long Term Change

Back in 2009, I did a series of posts covering the excellent book Crime Analysis For Problem Solvers. The book is published by the US DOJ's Problem Oriented Policing Center (POP Center). Because of the value I think this book has for crime analysts, and policing in general, I am going to re-post this series on here on the blog.

 This post in Crime Analysis For Problem Solvers is going to cover Step 26 - Take Account of Long Term Change. Law enforcement practice has come a long way. Agencies now regularly use sophisticated analysis tools to uncover trends and identify problems. Charts and graphs that were once only regularly found in the corporate world are now making regular appearances at police department staff meetings. It's also not uncommon to see crime analyst job postings that require college or even graduate level education in statistics. All this is because departments are applying business intelligence techniques to their workflow.

Time series analysis is measuring the number of crimes or other events over time. This step in Crime Analysis For Problem Solvers discusses three important ideas in time series analysis. They are:
  • Overall trend
  • Temporal cycles
  • Random fluctuations
All three of these concepts are important to understand when graphing crimes or events over time. If your graph of events over time exhibits change, you should attempt to identify the reason for these changes in order to better understand what is occurring. 

Overall Trend
This is the one your Chief will likely lose sleep over. Did your crimes increase or decrease this year? Overall trend, the authors state: "shows whether the problem is getting worse, better, or staying the same over a long period."

Temporal Cycles
Temporal cycles are those changes that occur regularly due to the rhythms found in normal life. A daily cycle might be the increase in traffic accident calls each workday around the morning and evening commutes. A weekly one may be an increase in DWI arrests on weekends. At my agency we have a yearly decrease in calls for service around January and February when the weather is the coldest.

Random Fluctuations
These are the hardest to explain. They may be caused by a one or two day increase in crimes committed by a lone offender or just minor changes in the "noise floor" of crimes. The authors describe the cause as "a large number of minor influences". 

The authors state the importance of understanding and using time series analysis comes from using it as a gauge of the effectiveness of your response to problematic crime.
Time series analysis is a powerful tool for evaluating the effectiveness of a response. The basic principle is to obtain a good idea of a problem's natural trends, cycles, and variation before the response is implemented, using the techniques just discussed. This tells you what you can expect from the problem in the future, if you did nothing about the problem. This provides a basis for examining time frames after the response. Changes in the trend, cycles, or even the random fluctuation suggest the response had an impact. The longer the time frames before and after, the greater the confidence you can have in your conclusions.
Next time we'll look at Step 27 - Know How To Use Rates and Denominators.

No comments:

Post a Comment

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.