Tracking Procedural Justice in processing arrestees: coding evidence from CCTV cameras in three custody suites

Research Institution / Organisation

University of Cambridge

In Collaboration With

Cambridgeshire Constabulary; Norfolk Police; Cambridge University

Principal Researcher

Kate Firman

Level of Research


Project Start Date

July 2020

Research Context

The researcher was the Head of Custody for Cambridgeshire before changing roles and was interested in tracking the extent to which custody officers display procedural justice (PJ) in their interactions with detainees during the booking-in process. This stage was chosen as studies have shown that PJ shapes reactions to decisions - in this case the decision to authorise detention as relayed by the custody suite officer. This research can then be used to identify whether there is a need to train custody staff in procedural justice and develop a bespoke training package for them.

Key Research Question and sub-questions:
How closely do custody suite encounters between arrestees and Custody Suite Officers (CSOs) match the procedural justice (PJ) standards for decision-makers treating persons who are subject to their authority, and to what degree does measurement of PJ displayed by CSOs vary across arrestees, CSOs and Custody Suites?

The study will test the following hypotheses (shaped from findings from other studies):

  1. Procedurally just encounters reduce overt resistance to police decisions

  2. Female officers are more likely to treat arrestees with PJ than their male counterparts

  3. Female arrestees are more likely to receive fair treatment than male arrestees

  4. Less experienced officers are more likely to treat arrestees with greater PJ than experienced officers

  5. Non-compliant citizens will be shown less PJ than compliant ones

  6. The level of PJ will vary across custody environments

Research Methodology

a. The unit of analysis is the arrestee-CSO encounter as captured on custody CCTV. The study defines the sampling frame from which to draw samples of CCTV video from the start to finish of an arrestee-CSO encounter as the time from which the arrestee is brought to the booking-in desk until the CSO has confirmed that they are authorising their detention and has completed the booking-in process.

b. The study draws from encounters involving 24 CSOs in Thorpe Wood and Parkside, two of which were female, and 11 CSOs in Kings Lynn, who also had two female CSOs. A random sample of cases was selected for Thorpe Wood and Parkside but all cases with a female CSO were selected for Kings Lynn. This was because of the low throughput of arrestees at Kings Lynn and the requirement to have an equal number of male / female CSO encounters in the study.

c. This study coded 150 encounters from arrests made in June, July and August 2020.

d. The majority of the coding was carried out by one person, an experienced police officer, who applied theoretical constructs and prior research in procedural justice to judge each encounter. A test re-test estimation of coding reliability of approximately 6% or ten encounters was carried out by another Mst student.

e. The coding instrument used by both coders was based on Jonathan-Zamir’s 2015 validated instrument for measuring PJ. Coding was carried out by using a pre-prepared checklist developed for this purpose. This allowed for any additional observations, including non-verbal and verbal communications to be recorded, making this study a mixed method design. Demographic characteristics of officers – for example, gender and length of service – and any use of force were coded. The average duration of the interaction between CSO and arrestee was 20 minutes and each one was reviewed and coded according to the four elements of PJ – voice, trustworthy motives, respect and neutrality. Each element was given a value, which has been combined to give an overall PJ score.

f. An equal number of cases were selected from each of the three custody suites used in the study to allow comparisons to be made.

g. There was a mix of offences for which the arrestees were detained which were scored using the Cambridge Crime Harm Index ((CCHI) for custodial and non-custodial offences.

h. The study will identify how PJ measured by CCTV coding relates to the use of force recorded in 30 of the cases and the level of compliance of an arrestee, recorded as non-compliant in 13 cases, by observing the CCTV footage and checking it against any use of force records. This includes the use of handcuffs at the booking-in desk.

i. What correlation, if any, is there between PJ scores and the fact of either the CSO or the arrestee being female compared to cases in which both CSO and arrestee are male?

j. What correlation, if any, is there between PJ scores and the fact the arrestee is not White British compared to cases in which the arrestee is White British?

k. What correlation, if any, is there between PJ scores and

  • CSO length of service in that role
  • Age at which the CSO was hired 
  • Total length of service in policing
  • Current age of CSO

Based on all the evidence above, what benefit might be gained from provision of training in PJ to CSOs, in terms of either PJ itself or more specific outcomes such as reducing assaults against officers in custody suites?

Units of analysis: 1) encounters between CSOs and arrestees; 2) CSOs with their average PJ scores on each dimension and overall; 3) Each custody suite as an average of CSO average scores
Data set size: 150 encounters
Time period: All CCTV records will be coded within 30 to 60 days after the recordings are made
Key measures of independent, dependent or descriptive variables: Similar to those used by Nawaz & Tankebe 2018
Research design: As indicated from answers to questions 2.a through 2.g.
Summary of Analytic Methods: Coding PJ concepts and aggregating scores by the major dimensions of PJ, then comparative analysis across encounters, officers, and custody suites

Interim reports and publications

​Not available

Date due for completion

March 2021
Return to Research Map