Journalist, Editor, Photographer, Writer
WHY DATA LITERACY MATTERS:
Violent Crime in Middle Georgia
This project is an inspection of the GBI's data on macon Georgia and its rising violent crime rate, and how data literacy affects our perceptions of rural and urban areas.
THE WORKSHOP: WHAT IS DATA LITERACY?
Data Literacy is a huge part of modern day society. News consumed by the public, whether it be printed, radio, or television; whether it be news or advertisement, uses statistics. To have a baseline understanding of a presented percentage, rate, or chart is just as important as understanding how to read.
To begin this “workshop” on data literacy, the first step is to look at the raw data, numbers in a dataset. The data in these examples is from the Georgia Bureau of Investigation (GBI), who tally up all the violent crime in Georgia as part of the Uniform Crime Reporting (UCR) program run by the Federal Bureau of Investigation (FBI). The population estimates used in the rate per capita calculations are from census.gov.
RAW NUMBERS: MURDER
When only looking at the raw instance data, it leaves an unbalanced view, there is no perspective to it. The GBI, therefore, uses more calculations after they collect the data, to apply it. These calculations help ensure the data is truly representative of what it is trying to show.
Special Agent in Charge of Region 13 Jerry Jones of the GBI explained that the Georgia Bureau focused especially on investigation, and their records were based on the incidents where they step in to help Georgia State Patrol, County Sheriffs and Municipal Police Departments.
“Bibb County is included in my jurisdiction along with 9 other counties.” Jones said. “The GBI’s approach to violent crime is providing investigative assistance to local law enforcement agencies. Being able to support each other, being here to provide investigative assistance and intelligence, information. We have a Georgia gang database, where we gather gang intel for the purpose of sharing it with local law enforcement agencies.”
When examining the data, the first question is whether Macon-Bibb, Dekalb and Fulton Counties can really be compared. If they all have different populations, that should affect their crime rates in vastly different ways, right?
RATE PER CAPITA: THE DIFFERENCE
Dr. André Waschka, a professor of statistics at Mercer University, explains that the rate per capita metric can be more helpful, but still is relative to itself when just treated as numbers.
“If you're taking the average of essentially two counties and one of them is lower and you're bringing that in, you start with one that has a higher murder rate per capita and you're bringing in a lower one. It's going to pull down your average as well.” Waschka said.
Waschka further explained that was where real statistical laws and analysis come in. When you combine sample sizes, they will even out, proven by the Law of Large numbers, assisted by Central Limit Theorem.
“The two main ones are the law of large numbers and central limit theorem,” Waschka said. “The fact that you have a non-normal distribution doesn't actually end up mattering because you are able to eventually use the fact that these means go to a normal distribution.”
The Law of Large Numbers, according to Statology, states, “as a sample size becomes larger, the sample mean gets closer to the expected value.” The more and more data that is collected, the closer the resultant mean gets to the true probability.
Therefore, even though some places in Georgia have both higher population and murder numbers, or a higher rate per capita despite lower population and numbers, the more the non-homogenous array of the population is measured, the closer the statewide statistics get to being an accurate representation of the state as a whole.
Central Limit Theorem, according to Boston University of Massachusetts, states, “if you have a population with mean and standard deviation, and take sufficiently large random samples from the population with replacement, then the distribution of the sample means will be approximately normally distributed.”
The theorem is essentially working backwards to prove the same end. The more samples you take, even out of a non-homogenous space like the state of Georgia, the closer you will get to the true representative mean.
With both of these concepts in hand, Georgia’s statistics can be understood, overcoming the original concerns. These two core concepts apply to statistics as a whole, and allow for a better understanding of data literacy, and instruct readers how to understand what data means in the Georgian, American and global societies that engage with it everyday.