Skip to main content

Table 4 Pairwise risk differences (Average Treatment Effect), Bonferroni adjusted across multiple treatments and comparison group, by estimating method

From: Estimating a treatment effect on recidivism for correctional multiple component treatment for people in prison with an alcohol use disorder in England

Estimating method

Risk Difference

Standard Error

Z statistic

Adjusted p-value

99% Confidence Interval

RA

 Control vs Pharmacological Treatment

−0.07

0.03

−2.33

0.118

− 0.147

0.007

 Control vs RNR

0.023

0.154

0.15

1

−.374

0.42

 Control vs PSI

− 0.005

0.017

−0.33

1

− 0.05

0.04

 Pharmacological Treatment vs RNR

0.093

0.157

0.59

1

−.0311

0.498

 Pharmacological Treatment vs PSI

0.064

0.034

1.86

0.377

−0.024

0.153

 RNR vs PSI

−0.029

0.155

0.19

1

−0.43

0.371

IPW

 Control vs Pharmacological Treatment

−0.095

0.037

−2.57

0.061

−0.190

0.0002

 Control vs RNR

0.094

0.045

2.07

0.231

−0.023

0.211

 Control vs PSI

0.012

0.018

0.68

1

−0.034

0.059

 Treatment vs RNR

0.188

0.058

3.24

0.0072*

0.039

0.339

 Pharmacological Treatment vs PSI

0.107

0.041

2.62

0.0528

0.002

0.213

 RNR vs PSI

−0.081

0.048

−1.68

0.558

−0.207

0.044

AIPW

 Control vs Pharmacological Treatment

−0.092

0.032

−2.83

0.0279*

−0.176

−0.008

 Control vs RNR

0.017

0.153

0.12

1

−0.377

0.412

 Control vs PSI

−0.015

0.018

−0.85

1

−0.062

0.031

 Pharmacological Treatment vs RNR

0.109

0.156

0.7

1

−0.294

0.513

 Pharmacological Treatment vs PSI

0.076

0.037

2.07

0.2307

−0.018

0.171

 RNR vs PSI

−0.033

0.154

0.21

1

−0.43

0.364

IPWRA

 Control vs Pharmacological Treatment

−0.079

0.029

−2.74

0.0369*

−0.154

−0.005

 Control vs RNR

−0.321

0.100

−3.19

0.0085*

−0.580

−0.062

 Control vs PSI

−0.015

0.018

−0.85

1

−0.061

0.031

 Pharmacological Treatment vs RNR

−0.241

0.104

−2.31

0.125

−0.510

0.028

 Pharmacological Treatment vs PSI

0.064

0.034

1.89

0.352

−0.023

0.151

 RNR vs PSI

0.305

0.102

2.99

0.0167*

0.042

0.568

  1. *p < 0.05;