College's evaluation of Day One assessment shows racial disparity
The College of Policing has published an evaluation of the Day One assessment centre pilot, which had a higher success rate among white candidates than BAME ones.Day One was developed to replace Police SEARCH which had been in place for 18 years. Its planned rollout nationally for 2020 has been delayed due to the pandemic.
It aimed to give new recruit candidates the experience of the challenges a new constable could face on their first day working in policing.
White candidates passed at almost twice the rate as black candidates (81.1 per cent compared to 44.3 per cent).
Four forces took part in the pilot which ran from April 2018 and November 2019: the Metropolitan Police, Thames Valley, Surrey and Sussex.
The exercises included early intervention, dispute settlement, stop and search, collaboration, interview, youth engagement and burglary. They were a mixture of role play, written exercises, responding to video clips with multiple choice answers and a collaboration exercise with another candidate. They were designed so that no prior knowledge of the role of a police officer or any policing knowledge is required in order to successfully complete them.
There were no exercises where large differences were found between the percentage score of BAME and white candidates. However, burglary is being reviewed by the college as it had the highest difference in percentage score.
Out of the 18,150 candidates, 30.5 per cent of them were BAME. They had a 47.7 per cent pass rate compared to 73.7 per cent for white candidates.
Overall, 22.1 per cent of the successful candidates were BAME. Up to 31 March 2019, there were 7.3 per cent BAME PCs nationally
In the report, the College said: “Ethnicity is explored in more detail, as this is the key factor (of all protected characteristics reported) which shows lower success rates for underrepresented character groups.”
70.8 per cent of candidates whose first language was not English failed, compared to 28.8 per cent who had English as a first language. Around a third of BAME candidates' first language was not English.
The College said it was worth highlighting that some underrepresented groups, such as women and those with a disability, had a higher success rate.
The process also favoured younger applicants. A higher proportion of BAME candidates were male and tended to be older.
Between the forces, there are “clear variations in overall success rate”. (Met 65 per cent, Thames Valley 76.3 per cent, Surrey 86.9 per cent and Sussex 97.1 per cent).
Of the participating forces, BAME candidates to join the Met had the lowest pass rate of all four at 47.4 per cent.
Based on research conducted by the college in 2019, it is likely a high proportion of the Met candidate pool do not have Functional Skills Level 2 in English Language, which the college say is the appropriate level for police recruitment.
The college also said it is likely that the difference in success rates is linked to differences in pre-assessment processes and different candidate pools across forces.
The Met did not apply any sifting process and all applicants attended the assessment.
The evidence suggests the gap in success rates between some BAME groups and the white group will be smaller for forces who carry out preselection or sifting of candidates: Surrey had a BAME candidate pass rate at 100 per cent and Sussex 97.1 per cent. Both went through a pre-selection process. Candidate numbers were low, but the college call the results for the two forces “encouraging”.
“The evidence suggests that implementing an appropriate recruitment strategy, including investment in attraction, sifting and positive action initiatives may be an effective means of tackling subgroup differentials at the Day One Assessment Centre.
“However, forces need to ensure that adverse impact is not being transferred to earlier stages of the recruitment process,” the report said.
The college said: “Commercial approval will be sought for the procurement of an appropriate national sift mechanism.”
Day One was due to be rolled out to all forces from the summer of 2020. However, due to the pandemic, all face-to-face officer recruitment has been postponed.
The college’s last analysis, the online assessment centre pass rate for white candidates was 76 per cent (2,461 out of 3,258) and for BAME 62 per cent (306 out of 496). A college spokesperson said: “A key difference with the online assessment process is that all candidates had passed either an in-force sift, or the College’s situational judgement test.”
A formal evaluation of the online assessment process will take place this autumn.
The college said they will review the evidence in relation to Day One, as well as the evaluation of the online assessment process, before setting out plans for the long term future of police assessment with a “fair and inclusive approach”. These plans are likely to incorporate elements of online and face-to-face assessment, the college said.
A spokesperson from the college said: “The College of Policing is working closely with policing leaders to ensure that the entire police recruitment system from end to end is fair and representative, whilst meeting the aims of greater diversity and high standards in police recruits.”
News Archive
- December 2023 (3)
- November 2023 (5)
- October 2023 (4)
- September 2023 (5)
- August 2023 (4)
- July 2023 (3)
- June 2023 (5)
- May 2023 (2)
- April 2023 (5)
- March 2023 (3)
- February 2023 (7)
- January 2023 (11)
- December 2022 (6)
- November 2022 (5)
- October 2022 (5)
- September 2022 (6)
- August 2022 (2)
- July 2022 (11)
- June 2022 (8)
- May 2022 (11)
- April 2022 (8)
- March 2022 (3)
- February 2022 (5)
- January 2022 (12)
- December 2021 (1)
- November 2021 (9)
- October 2021 (4)
- September 2021 (10)
- August 2021 (9)
- July 2021 (12)
- June 2021 (4)
- May 2021 (11)
- April 2021 (14)
- March 2021 (14)
- February 2021 (19)
- January 2021 (18)
- December 2020 (6)
- November 2020 (12)
- October 2020 (14)
- September 2020 (15)
- August 2020 (16)
- July 2020 (16)
- June 2020 (18)
- May 2020 (22)
- April 2020 (18)
- March 2020 (23)
- February 2020 (20)
- January 2020 (4)
- December 2019 (5)
- November 2019 (6)
- October 2019 (5)
- September 2019 (6)
- August 2019 (8)
- July 2019 (6)
- June 2019 (8)
- May 2019 (8)
- April 2019 (8)
- March 2019 (10)
- February 2019 (9)
- January 2019 (9)
- December 2018 (9)
- November 2018 (12)
- October 2018 (8)
- September 2018 (7)
- August 2018 (11)
- July 2018 (7)
- June 2018 (9)
- May 2018 (9)
- April 2018 (12)
- March 2018 (10)
- February 2018 (8)
- January 2018 (5)
- December 2017 (6)
- November 2017 (4)
- October 2017 (3)
- September 2017 (10)
- August 2017 (5)
- July 2017 (5)
- June 2017 (6)
- May 2017 (6)
- April 2017 (2)
- March 2017 (3)
- February 2017 (4)
- January 2017 (1)
- December 2016 (3)
- November 2016 (4)
- October 2016 (1)
- September 2016 (4)
- August 2016 (4)
- July 2016 (1)
- June 2016 (5)
- May 2016 (3)
- April 2016 (1)
- March 2016 (3)
- February 2016 (3)
- January 2016 (3)
- December 2015 (3)
- November 2015 (3)
- October 2015 (3)
- September 2015 (2)
- August 2015 (1)
- July 2015 (11)
- June 2015 (1)