- Copperas Cove teen arrested for high school bomb threat
- Killeen Police investigating storage unit burglaries
Chasing The Needle
This article is a repost from one I wrote for the blog The Crime Map. I figured with all the other posts about crime stats this week it was appropriate to revisit this.
When I was in the Navy, the helmsmen, the person who actually steered the ship, had to learn an important lesson. The ship's rudder controlled the direction the ship was traveling. The rudder was controlled by either a wheel or by a lever from the bridge of the ship. Near the helmsman's station was a compass repeater that indicated the direction the ship was traveling. Usually, the helmsman would receive an order to steer the ship a certain direction on the compass.
An experienced helmsman would often tell a new helmsman to avoid the temptation to "chase the needle". When the helmsman would turn the wheel, it might take a bit before the ship would begin to turn and the compass would then begin to show the desired course. If the helmsman was over anxious, they would wrongly assume that their rudder input was not having the desired effect and would overcompensate. When the ship then began to turn it would often overshoot the desired heading and lead to a back and forth fight between the helmsman over-steering and the ship slowly responding to this heavy handed input.
As a crime analyst I tell my Chief and his staff members this story often to help them avoid the desire to "chase the needle" over every variation in crime stats. Over the years, crime statistics have become more and more important in gauging the effectiveness of a law enforcement agency and by implication, the leadership of that agency. We've seen a number of studies that hint that this accountability is a major driving force in the reduction of crime we've seen in the past few years here in the U.S. An effect of this new accountability is that it is not uncommon for a police chief to be sacked if the department's crime numbers are high. This increase in accountability also leads to a tendency to overreact over apparent changes in the crime stats from month to month or even worse, week to week.
Understanding The Reason For Change
The DOJ funded Center for Problem Oriented Policing has a great book called Crime Analysis For Problem Solvers In 60 Small Steps. In the section Step 26 - Take Account of Long Term Change the authors look at time series analysis, or measuring events over time. This is what we do when we count the numbers of crimes or events and graph them over time. The authors state that there are three important factors to understand when conducting this type of analysis:
The Importance Of The Long Term
- Random Fluctuations
- Temporal Cycles
- Overall Trend
The first two, Random Fluctuations and Temporal Cycles are the probably the two factors least worthy of your Chief's anxiety. The authors of Crime Analysis For Problem Solvers describe Random Fluctuations as change "a large number of minor influences". Temporal Cycles are those daily, weekly or seasonal variations caused by the changing rhythms of normal cyclical patterns.
The Importance Of The Long Term
The last one, Overall Trend is the one that you should pay the most attention to. Generally, the overall trend is something that is measured over a longer period of time. This is where we answer what happened with our crime rate at the end of the year. These are the trends that are reported in the FBI's publication Crime In The United States which is published annually.
Thankfully, there are a number of tools to help us determine the Overall Trend. One technique involves graphing a three month moving average of month to month crime data to smooth out random fluctuations. Looking at successive time periods of data as discrete series in the same chart can help you to identify temporal cycles. Most spreadsheet programs can graph a trendline onto an existing chart to show you where your data is headed long term.
Separating minor temporal changes from the important ones will help you to know where you are actually headed and avoid "chasing the needle" with all the wasted effort and stress that goes along with it.
Separating minor temporal changes from the important ones will help you to know where you are actually headed and avoid "chasing the needle" with all the wasted effort and stress that goes along with it.