Skip to main content

Table 4 Logistic regression estimates and odds ratios predicting serious misconduct among police

From: A machine learning analysis of serious misconduct among Australian police

Variable β SE OR
Demographic variables
Secondary employment 1.56 0.09 4.38**
Awards or complimentary remarks 0.19 0.19 1.22**
Age 0.44 0.07 1.53*
Length of service 0.31 0.16 1.37*
Socio-economic status − 0.72 0.27 0.48*
Years at rank 0.29 0.09 1.39
Transfers − 0.34 0.26 0.73
Duty type: General duties 0.15 0.05 1.16
Rank: Senior Constable 0.10 0.29 1.10
Misconduct process variables
Unsubstantiated complaints 1.14 0.18 3.15**
Local Area Commander Warning Notice 0.44 0.15 1.55
Region Commander Warning Notice 0.49 0.72 1.64
Performance or conduct management plan 0.36 0.08 1.44
Management action 0.23 0.07 1.26
Misconduct types
Issues with an investigation 0.88 0.22 2.42
Improper use of force 0.64 0.11 2.00**
Drink driving 0.54 0.28 1.72*
Complaint related to court 0.59 0.47 1.79**
Breach code of conduct 0.59 0.72 1.80
Harassment, intimidation or bullying 0.49 0.08 1.64*
Allegation of corruption 0.22 0.20 1.25
Positive drug test 0.27 0.13 1.31
Disobey reasonable direction 0.37 0.33 1.44
Firearm related complaint 0.17 0.17 1.18
Untruthfulness/Dishonesty 0.16 0.29 1.17
Rudeness to members of the public − 0.09 0.02 0.91
Breach privacy, unauthorised disclosure of data − 0.45 0.21 0.64
Property or exhibits complaint − 0.59 0.55 0.55
  1. *P < 0.05
  2. **P < 0.01