logo
  • Donate
  • Database
    • Mass Shooter Database
    • Key Findings
    • Methodology
    • K-12 School Shooting Database
  • Our Book
  • Off-Ramp
  • More
    • About Us
    • Our Team
    • Our Work
      • By Jillian & James
      • Watch or Listen
      • Our Data on Social Media
      • TVP in the News
      • Media Mentions
      • Research Publications
      • Newsletter Archives
    • Contact Us
  • search

Methodology

Home > Methodology

Disclaimer: This project was supported by Award No. 2018-75-CX-0023, awarded by the National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. The opinions, findings, and conclusions or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect those of the Department of Justice.

The mass shooter database project included four phases:

  1. Creation of a comprehensive database of mass public shooters from 1966 to today coded on over 100 life history variables, including mental health history, trauma, interest in past shootings, and situational triggers.
  2. Examination of community-level socio-ecological factors of where mass public shootings take place, including, but not limited to, crime rates, measures of social inequality, community mobility, availability of mental health resources, and prevalence of gun stores.
  3. In-depth life history interviews with living mass shooters who are currently incarcerated and follow-up interviews with key stakeholders (e.g., family members, first responders, survivors, experts) in the communities where shootings took place.
  4. Dissemination of findings, creation of this public website, and implications for evidence-based prevention strategies.

The work is guided by rigorous ethical protocols from the Institutional Review Board at Hamline University and a steadfast commitment to no notoriety for mass shooters.

What is a Mass Shooting?

There are several gun violence problems in the United States, each demanding our attention and action, including:

  1. Suicide by firearm;
  2. Everyday gun violence (i.e., interpersonal conflicts and shootings related to gangs/groups or underlying crime);
  3. Domestic shootings;
  4. Unintentional shootings;
  5. Police-involved shootings;
  6. Mass public shootings.

Some problems are more prevalent and pressing than others, but each has its own victim and offender profile, underlying mechanisms, and potential solutions.

There is no universally accepted definition of a mass shooting. For a long time, mass shootings were studied as a variation on "mass killing", the generic term for multiple homicides that vary along what researchers agree are important dimensions: time, place, and method. For instance, someone who kills their victims in separate events is different from someone who kills them all at once. A person who kills in public is different from a person who kills in private (especially when private victims tend to be family members or intimate partners); and different still from a contract killer, bank robber, or gang member who kills in the commission of another crime. And an arsonist or bomber is different from a shooter.

Some sources define a mass shooting as any incident in which three or more victims were shot and killed, not including the perpetrator, while others use a threshold of four or more killed. Other sources define a mass shooting as any incident in which four or more people were shot or injured, which greatly increases the incident count.  Some sources include domestic violence, gang conflict, drug trade disputes, or robberies in their numbers. At The Violence Project, we focus on mass public shootings, defined by the Congressional Research Service as follows:

“a multiple homicide incident in which four or more victims are murdered with firearms—not including the offender(s)—within one event, and at least some of the murders occurred in a public location or locations in close geographical proximity (e.g., a workplace, school, restaurant, or other public settings), and the murders are not attributable to any other underlying criminal activity or commonplace circumstance (armed robbery, criminal competition, insurance fraud, argument, or romantic triangle).”

We acknowledge the limits of this definition. Every mass casualty event is a tragedy and many factors influence whether a threshold of four or more people killed is reached, including the accuracy of the shooter, the type and caliber of the firearm used, the number of rounds fired, the actions of first responders, proximity to the nearest hospital, and if/how many bullets hit vital organs. Any cut point is arbitrary, but this remains a widely agreed-upon standard. Further, the number of deaths is the strongest predictor of media coverage, which we use to help build our database.

By focusing only on mass public shootings, we exclude domestic homicides (although if 50% or more of victims are non-relatives killed in public then we include them). We also exclude mass shootings attributable to gangs or underlying criminal activity (e.g., a robbery or drug deal gone bad). A broader definition with a threshold of fewer deaths, non-fatal shootings, or any means/motive would certainly yield more cases. For examples, see:

Source Dates Casualties Location Exclusions
Violence Project 1966–

present

4+ killed

(excludes shooter)

Public When victims are exclusively family or friends;

shootings related to gangs or underlying crime

Gun Violence

Archive

2014–

present

4+ shot Any None
AP/USA Today/

Northeastern

University

2006–

present

4+ killed

(excludes shooter)

Any None
Everytown for Gun Safety 2009–

2018

4+ killed Any None
Stanford Mass Shootings of America 1966–

2016

3+ shot

(excludes shooter)

Any Shootings related to gangs, drugs, or organized crime
Washington Post 1966–

present

4+ killed

(excludes shooter)

Public When victims are exclusively family or friends;

shootings related to gangs or underlying crime

Mother Jones 1982–

present

3+ killed

(4+ killed ≤ 2013)

Public “Conventionally motivated crimes” such as gang violence and armed robbery

Our database was primarily built to explore the life histories of mass shooters, not simply track mass shootings. Our main unit of analysis is the shooter, not the shooting.  This means that in addition to the what, when, and where of mass public shootings, we study the why and how, plus who the victims of these crimes were, hopefully moving us closer to a world without mass shootings.

Building The Database

Data Sources Method Research Team Updates
Data Sources

To build the database, we used the following:

Primary Sources:

  • Written journals / "manifestos" / suicide notes etc.
  • Social media and blog posts
  • Audio and video recordings
  • Interview transcripts
  • Personal correspondence with perpetrators

Secondary Sources (all publicly available):

  • Media (television, newspapers, magazines)
  • Documentary films
  • Biographies
  • Monographs
  • Peer-reviewed journal articles
  • Court transcripts
  • Police reports
  • Medical records
  • School records
  • Autopsy reports

Community Variables:

  • U.S. Census Bureau
  • FBI Uniform Crime Reports
  • Google Maps


Method

Work began in Summer 2017. On October 1, a gunman on the 32nd floor of the Mandalay Bay resort in Las Vegas opened fire on a crowd of people at a country music festival, perpetrating the deadliest mass shooting in history. After that event, our undergraduate students began volunteering to work on the database — at first, for no pay or college credit, motivated solely by the need to do something, anything, to help stop the mass shooting epidemic. Over the next three years, they helped us code 180 mass shooters on nearly 200 different pieces of life history information.

Any coding is a subjective and interpretive process. Informed by existing datasets, the research literature, and frequently asked questions about mass shooters, the Principal Investigators generated a list of variables to be coded and created a codebook to define and detail how to code them. The codebook was then piloted on a small random sample of test cases and refined based on user-experience.

Once the codebook was finalized and coders were trained in its use, the database was populated as follows:

  1. Each mass shooter meeting the inclusion criteria (see definition above) was investigated twice by two separate coders, working independently from each other. 
  2. The two resulting datasets were then merged and compared.
  3. Any discrepancies were flagged and reconciled by consensus of the Principal Investigators, who did their own fact-checking and weighed the quality and quantity of the evidence, typically giving precedence to primary source material.
  4. The database was then divided up among the original coders and independently checked again.  
  5. Finally, the Database Manager conducted a full and final check, scrutinizing each and every cell.
  6. The Principal Investigators answered any queries resulting from the final check before approving publication. Responsibility for the contents of the database thus lies solely with the Principal Investigators. 

This is purposive sample and we present no comparison group. Comparisons are important. For example, 98% of mass shooters are men, but then 90+% of all homicide offenders are men. The majority of mass shooters are white, but the racial composition of the sample is comparable to the demographics of the U.S. population overall. 

Many of the factors correlated with mass shootings in the database are true of millions of people who never commit mass shootings. People may own guns, have traumatizing childhoods full of violence, reach a crisis point and want to die, think they’ve been victimized, even study other mass shooters, and still not commit a mass shooting. Personality and individual differences cannot be discounted. 

There is a low base rate of mass shooters and mass shootings are extreme and rare events—discrete occurrences of infrequently observed phenomena. For this reason, we caution against using the data for predictive modeling or cherry-picking one variable at a time to tell a particular story. For example, we see relatively high rates of mental illness among mass shooters—and rates of thought disorder that are considerably higher than those found in the general population. But this doesn’t mean mass shootings are exclusively caused by mental illness—the vast majority of people with mental disorders are never violent, and are more likely to be victims of violence than offenders. Classifications in the database are based on the available evidence, which sometimes includes demonstrated signs of undiagnosed mental illness and mental health evaluations conducted either before or after the perpetrators committed their attacks.

We have taken every step possible to find and verify sources and to rigorously fact-check the data, but the end result is not perfect. Data privacy laws (rightly) limit full access to official records. The source data were originally gathered for purposes different from our own. Media outlets have their own agendas and biases. Some cases are well reported on, others not so much, resulting in missing data. We may know more about recent cases, which reflects better reporting over time and more advocacy and awareness around of the topic of mass shootings. For these reasons, users should interpret trends over time with caution.

If you spot a mistake, please tell us so we can correct it. The full database is updated periodically and edited when new sources of information become available. New perpetrators are added annually.

Research Team

Principal Investigators:

  • Dr. Jillian Peterson
  • Dr. James Densley

Database Manager:

  • Stasia Higgins

Research Assistants (Versions 1–3):

  • Amanda Jensen
  • Kyle Knapp
  • Elliot Fay
  • Jessica Lindgren
  • Hannah Klumb
  • Hannah Peterson

Research Assistants (Versions 4 and 5):

  • Mary Dillon
  • Hannah Malicky
Updates

Version 6.1 (January 2022)

Fixed some classification errors in the firearms tab.

Version 6 (January 2022)

Added the 2021 shooting in Denver (case 183) and 2022 mass shootings at Uvalde, Tulsa, Highland Park, Raleigh, Colorado Springs, and Chesapeake

Added height and weight variables for mass shooters

Added mass shooting location counties, zip codes, and coordinates for geocoding

Added Fetal Alcohol Spectrum Disorder and Psychiatric Medication categories

Added large-capacity magazines to the firearms tab

Specified who exactly killed the shooter if they died at the scene

Version 5 (May 2022):

This was a major update. We added the November 2021 Oxford High School (case 181) and May 2022 Buffalo supermarket (case 182) shootings, then the entire codebook was updated with clearer and more detailed variable definitions.

Every case and code was revisited and changes were made based on new information available. Variables previously coded as “no” or “missing” were updated to “no evidence” which is a more accurate description for data drawn from public records and also eliminated the need for the old "Missing Data" tab.

Multiple columns for a single variable were collapsed into a single column containing multiple codes (i.e. domestic abuse, leakage, prejudice).

Hyperlinks for source materials were removed for ease of use are now available by request.

New Variables and Changes to Existing Variables:

Expanded "location" codes to include office buildings, post offices, and warehouses/factories and to better locate previously coded "workplace" shootings. A new "Workplace Shooting" variable was added for this reason.

Added a column to the “Victims” tab indicating the relationship between the victim and the shooter. Two new columns for “family member victim” and “romantic partner victim” were added to the full database. The old “domestic spillage” column was removed owing to redundancy with the new variables.

Removed "social alcohol" from the substance use and abuse variable.

Added specific psychiatric medications each shooter was prescribed, if known.

Removed the columns “Recent Breakup” and “Trouble at Work” incorporating them into the variable “Recent Stressor or Triggering Event”. The old “Recent Stressor” variable was then changed from a “yes/no” response to one that specified what kind of stressful or triggering event was occurring in the shooter’s life:

0 = None

1 = Recent break-up

2 = Employment-related stressor

3 = Economic stressor

4 = Family issue

5 = Legal issue

6 = Other

Re-coded "Adult Trauma" to specify what type of traumatic event occurred (rather than yes/no).

0 = none

1 = Death of a parent causing significant distress

2 = Death or loss of a child

3 = Death of a family member causing significant distress

4 = Trauma from war

5 = Traumatic accident

6 = Other

Re-coded "Childhood Trauma" from specifying the type of trauma to being a yes/no variable. The specifications were not very illuminating. Instead, we added columns specifying what Adverse Childhood Experiences (ACEs) each shooter had experienced. ACEs must be interpreted with caution because little information is available or known about the early childhoods of a majority of perpetrators in the database.

  • Physically abused
  • Sexually abused
  • Emotionally/verbally abused
  • Neglected
  • Witnessed violent treatment of mother
  • Parental substance abuse
  • Parental criminal record
  • Family member incarceration
  • Death of a parent in childhood

“History of Violence” was removed and the information is now captured in two new columns: “History of Physical Altercations” (which specifies if perpetrators had a history of getting into physical fights or directing violence at inanimate objects when angry at other people) and “History of Animal Abuse”.

The column “Criminal Record/Prior Police Contact” was separated into two columns: “Criminal Record” and “Known to law enforcement.” Previous coding conflated when police performed a mental health check on someone with an actual criminal record, which are now distinct.

The old “Criminal Record” category was expanded to capture what the FBI call “Part I” and “Part II” offenses that the perpetrator had been suspected, arrested, charged, or convicted of. Another new variable (“Highest Level of Criminal Justice Involvement”) now specifies the highest level of prior criminal justice involvement each perpetrator had. The old “previous homicides” variable was cut because this information is now captured in Part I crimes.

Part 1 Crimes:

0 = No

1 = Homicide

2 = Rape

3 = Robbery

4 = Aggravated assault

5 = Burglary

6 = Larceny-theft

7 = Motor vehicle theft

8 = Arson

Part II Crimes

0 = No

1 = Simple assault

2 = Fraud, forgery, embezzlement

3 = Stolen property

4 = Vandalism

5 = Weapons offenses

6 = prostitution or other sex offense (not rape)

7 = Drugs

8 = DUI

9 = Other

1 = Suspected

2 = Arrested

3 = Charged

4 = Convicted (includes plead guilty and sentenced to diversion)

The variable “Head Injury/Possible TBI” was added to the database.

In the “Religion” variable, “Atheist” and “No Religion” were collapsed into one code.

In the variables “Sexual Orientation”, LGB was updated to non-heterosexual.

Version 4 (July 2021)

We added five new mass shootings that occurred in the first half of 2021. We also updated 2019 and 2020 shooting cases based on new information available.

Version 3 (February 2021)

We added two new mass shootings from 2020 and the following new variables:

  1. Whether the shooter was an insider or outsider
  2. Voluntary or involuntary hospitalization
  3. Life expectancy of each victim based on birth year and demographics using CDC data
  4. Years of life lost for each victim based on life expectancy
  5. Minor fixes.

Version 2 (July 2020)

We released an updated and expanded version of the database with the following additions:

Cut 2 shooters who were found not guilty

Added one new shooting from 2019 and another from 2020 that met our definition

Updated small errors noted by users

Updated location so that “other” was  recoded to “outside” or “place of residence”

Expanded leakage

  • How plans were leaked (i.e. writing, in person, social media)
  • Who plans were leaded to (i.e. parent, friend, teacher, coworker, spouse)
  • Specific or unspecific plans

Expanded crisis 

  • Exact type of crisis prior to the shooting
  • Timeline of crisis
  • Recent or ongoing stressor
  • Notably depressed mood
  • Unusually happy or calm
  • Inability to perform daily tasks
  • Rapid mood swings
  • Increased agitation
  • Abusive behavior
  • Isolation
  • Losing touch with reality
  • Paranoia

New victim database - every person killed by a mass shooter in the database

  • Victim name
  • Victim age
  • Victim gender
  • Victim race
  • Victim relationship with perpetrator

New firearm database - each gun brought to a mass shooting coded separately

  • Exact make of gun
  • Type of gun
  • Was it used in the shooting?
  • Was it modified?
  • Extended magazine?
  • When was it obtained (month prior)
  • Legal purchase
  • Illegal purchase
  • Assembled with legal parts
  • Gifted
  • Theft
  • Caliber
  • Unknown

Role of psychosis in the shooting

  • 0 - psychotic symptoms played no role in the crime 
  • 1 - psychotic symptoms played a minor role in the crime (perpetrator experienced psychosis prior to the crime or during it, but it was not the main motivating factor)
  • 2 - psychotic symptoms played a moderate role in the crime (perpetrator was responding to delusions or hallucinations in planning or committing his or her crime, but they also also had other motives)
  • 3 - psychotic symptoms played a major role in the crime (perpetrator experienced psychosis prior AND during the crime, was responding to delusions or hallucinations in planning AND committing his or her crime, and had no other motives)

New violence variables

  • History of domestic violence
  • Previous sex offense
  • Use of violent video games
  • Gun proficiency

New mental health variables

  • Mental health treatment - mandatory or voluntary
  • Mental health treatment within six months of the shooting
  • Family history of mental illness

New family variables

  • Childhood socioeconomic status
  • Birth order
  • Number of siblings
  • Older siblings
  • Younger siblings

Other new variables

  • Performance in school
  • Community involvement
  • Hate group affiliation
  • Connection to fiction or pop culture
  • Armed person on the scene - specify who

Want to Know More About What We Do at the Violence Project?

LEARN MORE ABOUT THE VIOLENCE PROJECT

logo
Menu
  • Home
  • Donate
  • About Us
  • Mass Shooter Database
  • Contact
Important
  • Privacy Policy
  • Terms of Use
  • Site Map
Social media

© 2022 The Violence Project, All Rights Reserved.

Digital Marketing by
We use cookies on our website to give you the most relevant experience by remembering your preferences and repeat visits. By clicking “Accept”, you consent to the use of all the cookies.
We will not sell personal information.
Cookie SettingsAccept
Manage consent

Privacy Overview

This website uses cookies to improve your experience while you navigate through the website. Out of these, the cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these cookies. But opting out of some of these cookies may affect your browsing experience.
Necessary
Always Enabled
Necessary cookies are absolutely essential for the website to function properly. These cookies ensure basic functionalities and security features of the website, anonymously.
CookieDurationDescription
cookielawinfo-checkbox-analytics11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Analytics".
cookielawinfo-checkbox-functional11 monthsThe cookie is set by GDPR cookie consent to record the user consent for the cookies in the category "Functional".
cookielawinfo-checkbox-necessary11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookies is used to store the user consent for the cookies in the category "Necessary".
cookielawinfo-checkbox-others11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Other.
cookielawinfo-checkbox-performance11 monthsThis cookie is set by GDPR Cookie Consent plugin. The cookie is used to store the user consent for the cookies in the category "Performance".
viewed_cookie_policy11 monthsThe cookie is set by the GDPR Cookie Consent plugin and is used to store whether or not user has consented to the use of cookies. It does not store any personal data.
Functional
Functional cookies help to perform certain functionalities like sharing the content of the website on social media platforms, collect feedbacks, and other third-party features.
Performance
Performance cookies are used to understand and analyze the key performance indexes of the website which helps in delivering a better user experience for the visitors.
Analytics
Analytical cookies are used to understand how visitors interact with the website. These cookies help provide information on metrics the number of visitors, bounce rate, traffic source, etc.
Advertisement
Advertisement cookies are used to provide visitors with relevant ads and marketing campaigns. These cookies track visitors across websites and collect information to provide customized ads.
Others
Other uncategorized cookies are those that are being analyzed and have not been classified into a category as yet.
SAVE & ACCEPT