Frequently Asked Questions
1) Can we provide the tally of absolute data for each metric? For instance, how many statements overall? How many specific statements by metric?
All metrics are “absolute data for each metric.”
Hours of video with speech are the hours of video content with actual speech content that can be transcribed. The number of videos from the same period were 2,673 for 2021 and 3,500 for 2022.
The metrics of count of “explanation,” “non-compliance,” and “escalation (force)” are the count of actual statements that were labeled as “explanation,” “non-compliance,” and “escalation (force).” A statement is approximately 5 to 20 seconds in length, that indicate “explanation,” “non-compliance,” and/or “escalation (force)” occurred in a particular incident. Therefore, these counts are “specific statements by metric.” Count of “officer unprofessional behavior” or “civilian rudeness” are based on shorter phrases (“f*** you” would be one count and if it’s repeated twice, it would be two counts). Unlike the other categories, these are measured in at least 2 second in length.
Therefore, these counts are “specific by metric.”
Total
Videos
Speech
Duration
Duration of
Explanation
Duration of
Non-
Compliance
Duration of
Officer
Unprofessional
Behavior
Duration
of Civilian
Rudeness
Duration of
Escalation
(Force)
2021 25,116 101.8 12.4 15.2 51.8 50.2
2022 28,500 152 14.2 13.9 85.8 41.8
*Duration in minutes
2) Did anyone authorize Truleo to prepare a case study on APD? If yes, who? Yes. Chief Joshi made the request to prepare the case study in order to assess efficacy of the software and also authorized this.
3) Has Truleo removed any City of Alameda references from their web page? While the Alameda case study was initially used by Truleo to market its services, staff requested that they no longer actively market their services using Alameda’s case study and staff has requested that they remove the case study from their webpage. The webpage does still include high-level summary data from the case study. Although Truleo received permission from the Chief of Police to assess performance of the technology and prepare the case study based on APD data, Truleo will get explicit approval from the City Manager before using the case study going forward.
4) What is the quantitative data that supports Truleo’s findings in their case study that are statistically different?
This analysis was looking at the entire population of data. Therefore, a sampling error/margin of error is not applicable (or rather, effectively 0%). However, observational error and measurement error may be factors.
Notably, one could argue that the same effects could be seen if officers turned their cameras on more often for positive events/encounters and less often for negative events/encounters between 2021 and 2022. However, there are two factors that make that conclusion unlikely:
a. Camera activations increased from 2021 to 2022
b. Reported use of force from APD matched the trend observed over the same period of time.
5) Has anyone from City staff analyzed Truleo’s data findings? Yes. APD reviewed Truleo’s findings and found similar findings. While Truleo’s software captures language that indicates an escalation and possible use of force, APD counts actual use of force incidents. Even still, APD found the reductions to be similar as indicated in Truleo’s case study.
One of the purposes of Truleo’s technology is to flag language that contain escalations.
These flags allow supervisors and/or commanders to assess if force actually occurred.
At a minimum, these flags afford APD the opportunity to conduct their own review of escalations. If an escalation results in the application of force, a separate Use of Force report is completed. Use of Force reports confirm force actually occurred and provide an assessment on the decision to use force, tactics used, policy compliance, and an overall review.
Although the Truleo case study showed a 36% reduction in language indicative of force, APD’s own analysis found a 50% reduction in actual Use of Force reports corroborating Truleo analysis.
6) Who prepared responses to these questions?
Dr. Shastry, PhD.
7) Are there other agencies that use this technology?
Yes. Alameda was the first to incorporate this technology. Many agencies across the country are using this technology, such as Ann Arbor, Michigan, Aurora, Colorado, and Anaheim, California.
8) Are there agencies that stopped using this technology? Seattle and Vallejo Police Department. Both agencies received objections from their unions.
Seattle PD believed management could “…use the footage they collect and store for many purposes, including surveillance.” (Axios, 2023). Additionally, the Seattle Police Union believed this technology was “…a form of spying on officers.” (Axios, 2023).
Seattle Police Officer Guild President Mike Sloan said, “The Department was spying on their employees without letting them know.”
The
Vallejo PD ended its agreement with Truleo after the Vallejo Police Union “…argued it should have been consulted first.” (Open Vallejo, 2023).
9) Has anyone from APD privately invested in Truleo? Is anyone from APD hired to help Truleo sell their products?
No. Such action would require compliance with Form 700.