Skip to main content

Table 5 Empirical standard error for the number of observations each arm = 100

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

Distribution of productivity loss outcome Scales in the two armsa OLS NB ZTNB ZG Three-part
x = mean
 80:15:5/60:30:10 Equal Scale 2.708 (0.027) 2.813 (0.028) 2.714 (0.027) 2.716 (0.027) 2.677 (0.027)
 80:15:5/60:30:10 Unequal Scale 2.756 (0.028) 2.867 (0.029) 2.766 (0.028) 2.768 (0.028) 2.718 (0.027)
 60:35:5/40:50:10 Equal Scale 2.734 (0.027) 2.793 (0.028) 2.750 (0.028) 2.752 (0.028) 2.671 (0.027)
 60:35:5/40:50:10 Unequal Scale 2.744 (0.027) 2.806 (0.028) 2.761 (0.028) 2.763 (0.028) 2.689 (0.027)
 50:40:10/30:55:15 Equal Scale 3.025 (0.030) 3.096 (0.031) 3.056 (0.031) 3.058 (0.031) 2.981 (0.030)
 50:40:10/30:55:15 Unequal Scale 3.051 (0.031) 3.119 (0.031) 3.082 (0.031) 3.084 (0.031) 3.013 (0.030)
x = 0
 80:15:5/60:30:10 Equal Scale 2.708 (0.027) 2.447 (0.024) 2.414 (0.024) 2.416 (0.024) 2.415 (0.024)
 80:15:5/60:30:10 Unequal Scale 2.756 (0.028) 2.499 (0.025) 2.465 (0.025) 2.467 (0.025) 2.456 (0.025)
 60:35:5/40:50:10 Equal Scale 2.734 (0.027) 2.448 (0.024) 2.495 (0.025) 2.498 (0.025) 2.504 (0.025)
 60:35:5/40:50:10 Unequal Scale 2.744 (0.027) 2.458 (0.025) 2.489 (0.025) 2.492 (0.025) 2.497 (0.025)
 50:40:10/30:55:15 Equal Scale 3.025 (0.030) 2.711 (0.027) 2.779 (0.028) 2.782 (0.028) 2.781 (0.028)
 50:40:10/30:55:15 Unequal Scale 3.051 (0.031) 2.746 (0.027) 2.818 (0.028) 2.822 (0.028) 2.827 (0.028)
x = 30
 80:15:5/60:30:10 Equal Scale 2.708 (0.027) 4.118 (0.041) 3.346 (0.033) 3.348 (0.033) 3.234 (0.032)
 80:15:5/60:30:10 Unequal Scale 2.756 (0.028) 4.100 (0.041) 3.388 (0.034) 3.390 (0.034) 3.266 (0.033)
 60:35:5/40:50:10 Equal Scale 2.734 (0.027) 3.613 (0.036) 3.228 (0.032) 3.227 (0.032) 3.019 (0.030)
 60:35:5/40:50:10 Unequal Scale 2.744 (0.027) 3.623 (0.036) 3.252 (0.033) 3.250 (0.032) 3.060 (0.031)
 50:40:10/30:55:15 Equal Scale 3.025 (0.030) 3.847 (0.038) 3.518 (0.035) 3.516 (0.035) 3.265 (0.033)
 50:40:10/30:55:15 Unequal Scale 3.051 (0.031) 3.841 (0.038) 3.523 (0.035) 3.520 (0.035) 3.281 (0.033)
  1. OLS ordinary least squares, NB negative binomial, ZTNB two-part model – logistic regression for the probability of being zero, and generalized linear regression with zero-truncated NB distribution for the non-zeros, ZG two-part model – logistic regression for the probability of being zero, and generalized linear regression with Gamma distribution for the non-zeros; Three-part: multinomial logistic regression for the probabilities of being zero and 60 and generalized linear regression with Beta distribution for the those with values in (0, 60) (transformed to (0, 1))
  2. afor truncated negative binomial distributions of productivity loss outcomes in the two arms