Skip to main content

Table 2 Simulation results.

From: Analysis of an ordinal outcome in a multicentric randomized controlled trial: application to a 3- arm anti- malarial drug trial in Cameroon

   

Methods

  

Bias (*)

  

1-β(α)

 

Parameters

Real values

ML

IGLS

GEE

ML

IGLS

GEE

ML

IGLS

GEE

Cut-points

          

α 1

-2

-2.149

-2.149

-2.143

0.014

0.007

0.0137

   
  

(0.236)

(0.237)

(0.237)

(7.34)

(1.98)

(6.76)

   

α 1

-1

-1.053

-1.053

-1.040

0.012

0.002

0.0136

   
  

(0.214)

(0.211)

(0.217)

(7.35)

(1.69)

(7.91)

   

α 3

0

-0.106

-0.106

-0.103

0.0055

-0.0043

0.0025

   
  

(0.206)

(0.204)

(0.207)

(3.57)

(1.6)

(1.56)

   

Treatment effects

γ 01

-1

-1.016

-1.016

-0.941

0.0077

0.0195

0.0098

0.938

0.948

0.94

  

(0.242)

(0.240)

(0.248)

(3.22)

(2.42)

(3.75)

(0.014)

(0.013)

(0.015)

γ 02

1

1.078

1.078

1.082

-0.005

-0.0043

-0.0111

0.943

0.943

0.945

  

(0.214)

(0.215)

(0.217)

(2.27)

(2.20)

(5.05)

(0.003)

(0.001)

(0.001)

Center effect

          

C1

 

0.218

0.218

0.168

      
  

(0.224)

(0.224)

(0.228)

      

C2

 

-0.014

-0.014

0.013

      
  

(0.228)

(0.226)

(0.230)

      
  1. Fixed models. Column 2 corresponds to the real values used for simulating data.
  2. Columns 3-5 correspond to the mean estimates according to different methods with the corresponding standard error estimates in brackets; columns 6-8 correspond to the corresponding absolute and relative biases (*); columns 9-11 corresponds to the coverage probability and, in brackets, the type I error for the treatment effect only.
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us