Skip to main content

Table 2 Mean bias and imprecision of individual scores for depression, n = 52 for random, demographic and subscale dependent missingness mechanisms

From: Handling missing items in the Hospital Anxiety and Depression Scale (HADS): a simulation study

Method Bias Imprecision n methoda
Random Demog Subscale Random Demog Subscale
25 % missing item rate, psub = 0.5 pitem = 0.5
 Subject mean 0.083 0.002 −0.060 1.638 1.707 2.177 52
 Subscale mean 0.004 −0.007 −0.010 2.316 2.533 3.483 52
 Subscale ½ mean 0.006 −0.002 −0.003 0.617 0.637 0.886 40
 Item mean 0.001 −0.360 −1.110 2.755 3.453 4.745 52
 Multiple imputation 0.016 −0.087 −0.324 1.727 2.043 2.623 52
 Multiple imputation ½ 0.010 −0.027 −0.119 0.635 0.738 0.965 40
10 % missing item rate, psub = 0.5 pitem = 0.2
 Subject mean 0.028 0.002 −0.025 0.459 0.474 0.632 52
 Subscale mean −0.003 0.001 −0.002 0.526 0.543 0.767 52
 Subscale ½ mean −0.002 0.001 −0.000 0.437 0.453 0.631 47
 Item mean −0.002 −0.123 −0.372 0.747 0.897 1.052 52
 Multiple imputation 0.005 −0.012 −0.075 0.532 0.612 0.773 52
 Multiple imputation ½ 0.001 −0.007 −0.063 0.472 0.545 0.672 47
10 % missing item rate, psub = 0.2 pitem = 0.5
 Subject mean 0.034 −0.003 −0.082 0.674 0.685 0.991 52
 Subscale mean −0.001 −0.004 0.001 0.986 1.025 1.468 52
 Subscale ½ mean 0.004 −0.002 −0.001 0.212 0.212 0.313 47
 Item mean −0.002 −0.132 −0.702 1.102 1.310 3.435 52
 Multiple imputation 0.005 −0.027 −0.234 0.662 0.751 1.276 52
 Multiple imputation ½ 0.002 −0.007 −0.074 0.195 0.222 0.363 47
2 % missing item rate, psub = 0.1 pitem = 0.2
 Subject mean 0.006 −0.000 −0.023 0.091 0.098 0.151 52
 Subscale mean −0.000 −0.000 0.001 0.106 0.110 0.158 52
 Subscale ½ mean 0.000 0.000 0.001 0.083 0.086 0.128 51
 Item mean 0.000 −0.029 −0.164 0.141 0.177 0.541 52
 Multiple imputation 0.001 −0.004 −0.038 0.101 0.112 0.192 52
 Multiple imputation ½ 0.000 −0.002 −0.033 0.086 0.193 0.163 51
  1. 1000 simulated datasets
  2. aSample size after imputation