Does Population Size Predict Violence?
During a recent discussion on gun control, an individual made a seemingly reasonable assumption that crime rate was influenced by population size. I didn't disagree with that assumption and my gut told me that it was probably true. However, I decided I needed more data before I could come to that conclusion.
The first thing I did was narrow down what type of behavior I wanted to look at. In this case, I wanted to look at Violent crimes. This is violent crimes as defined by the FBI:
In the FBI’s Uniform Crime Reporting (UCR) Program, violent crime is composed of four offenses: murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault. Violent crimes are defined in the UCR Program as those offenses which involve force or threat of force.
Source: FBI Website
So now that I knew what I wanted to look at I needed data to analyze. I found the following data on another FBI Page.
This data had to be cleaned up to only include population and total number of violent crimes per state.
|State||Population||# of Violent Crimes|
|District of Columbia4||617,996||7,429|
Next, I copied this data over to SPSS which is a data analysis software. For those of you unfamiliar with SPSS and statistical analysis in general, think of it as software that helps you draw conclusions based off mathematical formulas. These analysis help you make probability statements (similar to weather predictions) to determine if there are relationships or differences.
I will be helping you (the reader) determine what we can conclude from the SPSS outputs.
Before I ran the analysis I was aware of certain limitations with these data. First, there are many things not being taken into account. To begin with, we aren't taking into account how many cities are in each state, the rate of poverty per state, amount of land, or other variables which could easily moderate any data and change the conclusions we draw from this given data set. Basically, we are working within limitations.
On to the analysis...
ANOVA Stands for Analysis of Variance. The important information from this table is the column all the way to the left under Sig. This stands for "significance". Basically what you can interpret from the .000 is that there is less than .01% chance that the population and number of violent crimes are not related.
Adjusted R Square is the value we want to look at in this table. Roughly what this translates to is that about 95% of violent crimes can be predicted by a change in population. When population increases, so too does the number of violent crimes. However, this does NOT mean that population increase CAUSES increases in violent crimes, only that you can predict an increase (with a very high degree of certainty in this case) in violent crimes.
I intend to analyze some more data soon. For example, I'd like to see how states with strict gun laws compare to states with less strict gun laws. This will take some time, stay tuned.