COLLECT DATA
Our guidelines for collecting data will help you understand which risk factors are causing disparities.
Strengthening data collection will also shed light on which of those risk factors are driven by policies and behaviors, rather than things beyond a department’s control (such as poverty or crime rates). These standards can also help ensure that departments participating in the Justice Navigator collect the data required to receive all possible analyses on stops and use of force. CPE guidance for collecting data on calls for service and officer-initiated activity in computer aided dispatch (CAD) systems is forthcoming.
How to collect data for stops
1
Record all stops
Every single one. Every single time. Stop records should clearly distinguish vehicle stops from pedestrian stops. If the department collects data on consent-based encounters such as witness interviews, these should be clearly distinguished from stops in which the person is not free to leave. Avoid using terms like “field interview” that could apply to either a stop or a consent-based encounter.
2
Record all stop dispositions
Including traffic citation, citation in lieu of arrest, arrest, or warning.
3
Record the perceived racial group, age, and gender of the person in every stop incident
Train officers not to ask people to self-identify their racial group, but to record their own perception of the person’s racial group. Officers should not be able to enter “unknown” for this category, and should only be able to enter “other” if also providing a description of what “other” means.
4
Record the detailed location of every stop
Use latitude and longitude coordinates; a complete street address, block number and street name; the closest intersection; or closest highway and highway exit.
5
Record the reason for each stop
If the reason for a stop is a traffic violation or suspected criminal activity, the stop record should indicate the vehicle or criminal code that the person was suspected of violating. Stop records should indicate whether the stop was intelligence-led.
6
Record every search conducted during a stop
Records of searches should specify the type of search, the reason for the search, whether or not the person consented to the search, and the search outcome, including whether and what contraband was found.
How to collect use of force data
1
Record all use of force incidents according to a clear definition of reportable force
Officers should report any force witnessed or used beyond hand control or escort techniques (i.e., handcuffing) of a complaint person. Force reporting should include the use of any tool, weapon, technique, or body part to overcome resistance or ensure compliance, regardless of injury or complaint of injury. The data should be able to distinguish which officer used which type of force on which person.
2
Record the perceived racial group, age, and gender of the person in every use of force incident
Train officers not to ask people to self-identify their racial group, but to record their own perception of the person’s racial group. Officers should not be able to enter “unknown” for this category, and should only be able to enter “other” if also providing a description of what “other” means.
3
Record the detailed location of every use of force incident
Use latitude and longitude coordinates; a complete street address, block number and street name, the closest intersection; or closest highway and highway exit.
4
Record whether de-escalation techniques were used
5
Record all types of force used in the incident
Including police weapons or tools used.
6
Record the nature of contact preceding the use of force incident
Such as officer-initiated or a 911 call.
7
Record the nature of any resistance encountered
8
Record whether the person was arrested and the nature of the offense for which they were arrested
This may be distinct from the call type.
9
Record whether any officer or the person subjected to force in the incident was injured, was hospitalized or taken to the hospital, or died
General guidelines for collecting data
1
Record officer information
In all stop and use of force records. Department data collection protocols should generally provide for the inclusion of the work unit assignment (at the time of the incident), racial group, gender, date of hire, as well as the unique identifier of each officer involved in a stop or use of force incident (learn about how we protect identifying information here).
2
Store data in a spreadsheet or electronic database
The data elements listed above must be available in a consistent format, and stored in a spreadsheet or database, in order to be analyzed. Information that is stored only on paper, in narrative form, or in audio or video format is essentially useless for quantitative analysis. Converting such data into an analyzable form is typically extremely burdensome and time-intensive. Collecting narrative data about contacts with the public is important for legal purposes and facilitating supervisor review, but is not a substitute for the uniform collection of data stored in a spreadsheet or database.
3
Implement procedures to ensure data integrity
Departments should train all officers on when and how to collect data on stops and use of force. Use of force reporting forms should align with and clearly delineate which actions the department counts as force and expects to be reported as a use of force incident. Supervisors should regularly review records to ensure that officers are completing them properly. Departments should have clear procedures in place to audit data and to immediately address any problematic findings.
How these standards help departments receive Justice Navigator results
Following the above guidelines will help departments receive the full set of analyses available in Justice Navigator assessments. A list of specific variables required for each individual analysis in the assessment is available in the “More information” section beneath each chart. Generally speaking, CPE requires valid data for at least 90% of observations for each required variable and for data to be reasonably complete and consistent for a period of at least three years.