New variables in K-12 School Shooting Database v4.1

Was a shooting targeted or random? Did it involve students or just happen on school property? Was it in a single room or multiple rooms inside the school?

Any researcher will tell you, it’s much easier to sort existing data than to collect and code it yourself. The K-12 School Shooting Database is a relational database with different tabs for the details about the incident, shooter(s), victim(s), and weapon(s). Each incident has a unique ID number that connects to the shooter, victim, and weapon.

To make the research process easier for everyone who uses this data, I’ve added new variables to the +3,000 cases including:

• GV_Type: Active Shooter/Planned Attack, Spontaneous, Targeted

• Involves_Students: Yes, No

• Targets: Random Shooting, Victims Targeted, Neither, Both

• Multiple_Locations: Single Room, Multiple Rooms, Hallway, Outside

• Preplanned: Yes, No

• SRO_School: Yes, No

• Security_Screening: None, Unknown, Metal Detectors, Manual Search, Armed Guards, Unarmed Guards

• Screening_Outcome: Outside/Off-Property, Bypassed, Failed

• Shots_Fired: Value (99 = more than 1 shot fired, unknown total)

• School Lockdown: Yes, No

• Campus_Type: Urban Street Highrise, Urban Street Standalone, Urban with Mini Campus, Single Building Indoor Hallways, Multiple Attached Building, Multiple Detached Building, Individual Classroom Buildings, Mega-School (see: There are 9 types of school campuses. Each needs a different security strategy.)

To request an updated excel copy of the raw data, email k12ssdb@gmail.com with your full name, institution/organization, contact information, how you plan to use the data, and a statement affirming that your use of the data will be cited.

Uses of New Variables

These new variables can be analyzed on their own, combined with other variables in the dataset, or compared to other data sources. They have real world applicability to school security.

For example, many schools are fortifying buildings by having a “single point of entry” or secure entry vestibules. Looking at the incidents that are coded for Campus_Type, the most frequent location for a shooting is a campus with multiple detached buildings

On a campus with multiple buildings (many have dozens of individual classroom buildings), it’s practically impossible to have a secure vestibule going into each building. Even if the inside of each building is secure, students still need to walk between them each time they change classes.

Another variable with real world application is Multiple_Location. There is a false assumption that school shooters go into multiple rooms looking for as many victims as possible. In reality, most shootings are either outside, begin/end in a single room, or take place in the open hallways.

It’s very rare for a shooting to take place inside multiple different rooms of the school. Since this data goes back to the 1960s, most of these incidents took place before classroom doors were locked and fortified. Even when classroom doors were either open or unlocked, school shooters usually don’t enter multiple rooms.

The new police report on the Covenant School shooting in Nashville reinforces this finding as the shooter walked around the building for 10 minutes without entering unlocked classrooms that had victims inside. From looking at the CCTV footage from the CVPA High shooting in St. Louis, the shooter also stayed in the hallways while he waited for police to find him.

I used the broadly inclusive definition that a school shooting is: all shootings at schools includes when a gun is fired, brandished with intent to harm, or a bullet hits school property for any reason, regardless of the number of victims, time, or day of the week. One of the most common criticism is: there are a bunch of gang shootings in the middle of the night that drive up the number.

This claim is objectively false. Just over 80% of the shootings that I’ve recorded involve students and staff at the school.

Keep the Research Rolling

I’m looking forward to getting emails requesting this new version of the database. Google Scholar is currently showing 346 papers citing the K-12 School Shooting Database.

I hope the number of citations can keep growing with even more interesting and novel findings that can provide empirical evidence for understanding this problem, preventing shootings before they happen, and improving school safety.

Speaking of empirical findings, if you missed it be sure to checkout my conversation with Dr. Jens Ludwig about his new book 'Unforgiving Places'. Why do two neighborhoods in Chicago have dramatically different rates of gun violence? One of them is an 'unforgiving place' to live.

---

David Riedman is the creator of the K-12 School Shooting Database, Chief Data Officer at a global risk management firm, and a tenure-track professor. Listen to my weekly podcast—Back to School Shootings—or my recent interviews on Freakonomics Radio, New England Journal of Medicine, and my article on CNN about AI and school security.