I drive by an elementary school on my way to work every day. More than once there’s been a police cruiser idling in the school’s parking lot with lights flashing and the officer standing nearby.

Although those incidents never involved a major crime, on several occasions this year the Tulsa World has chronicled arrests at schools.

In January, police arrested an 18-year-old man found with a stun gun, two samurai swords and six knives in his car in a high school parking lot.

In February, police arrested a 59-year-old man after he pointed a gun at students standing outside a Tulsa high school.

And later that month, police arrested a Tulsa County music teacher for allegedly fondling a 17-year-old male student.

There were others incidents, most of which were minor. Some included fights, vandalism and drug use.

Not long after the teacher’s arrest, I called my contact at the Tulsa Police Department’s information technology department and requested a database of all calls from area schools to police since 2005.

I wanted to see what types of crimes happened on campus. What schools were particularly violent? Has crime increased in recent years? How did elementary schools, such as the one I pass every day, stack up against their high school counterparts?

At the suggestion of the World’s education reporter, I also requested similar data from the police department in Broken Arrow, a city of about 91,000 just outside Tulsa.

It could be interesting to compare crime within each city’s schools, she said. Broken Arrow, unlike Tulsa, does not carry a big-city reputation and the negative connotations that typically follow. Despite its growing population, Broken Arrow is still known as a quiet, peaceful suburb.

Taming the call list
Data from both departments were free and came surprisingly fast. Within a week I received a Microsoft Excel spreadsheet from Tulsa and Broken Arrow police, but I immediately noticed a problem with my request.

Since 2005, Tulsa and Broken Arrow schools combined logged more than 18,500 calls to police. That was a red flag.

Although the calls from schools to police would be helpful, the data I really wanted to analyze were calls that resulted in a charge or an arrest.

That data would provide a more accurate description of crime in area schools. It would eliminate most calls for traffic stops, stray animals, abandoned vehicles and the like. It would also eliminate calls where police found no evidence of wrongdoing.

Gavin Off is the data editor at the Tulsa World in Oklahoma and a former analyst for the IRE and NICAR Database Library.