Skip to main content

Table 1 Parameters in the simulation study

From: How to analyze work productivity loss due to health problems in randomized controlled trials? A simulation study

Parameters 80:15:5/60:30:10 60:35:5/40:50:10 50:40:10/30:55:15
Number of observations each arm: Nobs 50; 100; 200 50; 100; 200 50; 100; 200
Multinomial distribution of productivity loss outcomes
 Probability of zero loss: P1(arm, 0)
  Arm 1 0.80 0.60 0.50
  Arm 0 0.60 0.40 0.30
 Probability of loss between 0 and Max: P2(arm, 0)
  Arm 1 0.15 0.35 0.40
  Arm 0 0.30 0.50 0.55
 Probability of max loss: P3(arm, 0)
  Arm 1 0.05 0.05 0.10
  Arm 0 0.10 0.10 0.15
 Coefficient of x
  γ1 −0.02 −0.02 −0.02
  γ2 −0.01 −0.01 − 0.01
Truncated negative binomial distribution of productivity loss outcomes between 0 and Max
 Mean: E for Arm 1 15.00 15.00 15.00
 Mean: E for Arm 0 20.00 20.00 20.00
Max loss 60.00 60.00 60.00
 Equal scale
  SD for Arm 1 13.34 13.34 13.34
  SD for Arm 0 14.63 14.63 14.63
 Unequal scale
  SD for Arm 1 12.00 12.00 12.00
  SD for Arm 0 16.00 16.00 16.00
 Coefficient of x: a 0.15 0.15 0.15
Baseline productivity loss
 Probability of zero loss: P 0.30 0.30 0.30
 Mean for non-zero loss: μx 20.00 20.00 20.00
 SD for non-zero loss: sdx 16.00 16.00 16.00
  1. SD standard deviation