Skip to main content

Advertisement

Table 6 Marginal effects for population average models

From: Is the value of a life or life-year saved context specific? Further evidence from a discrete choice experiment

Predictor ∂ UB - UA/∂ xj SE 95%CI xj ∂ UB - UA/∂ xj SE 95%CI xj
  Lives saved Life-years saved
Cure(B – A)~ -0.2118 0.028 (-0.27,-0.16) 0 -0.2082 0.026 (-0.26,-0.16) 0
AgeGrp1(B – A) 0.3222 0.037 (0.25, 0.39) 0 0.1862 0.036 (0.12, 0.26) 0
AgeGrp2(B – A) 0.1484 0.034 (0.08, 0.22) 0 0.0750 0.033 (0.01, 0.14) 0
AgeGrp4(B – A) -0.0952 0.028 (-0.15,-0.04) 0 0.0047 0.032 (-0.06,0.07) 0
Evidence(B – A)~ 0.1714 0.023 (0.13, 0.22) 0 0.1643 0.023 (0.12, 0.21) 0
Fault(B – A)~ -0.1455 0.024 (-0.19,-0.10) 0 -0.1640 0.026 (-0.21,-0.11) 0
$Private(B – A)^ -0.0014 0.001 (-0.00,-0.00) 0 -0.0019 0.001 (-0.00,-0.00) 0
Effect(B – A) 0.0085 0.001 (0.01, 0.01) 0 0.0002 0.000 (0.00, 0.00) 0
$Cost(B – A)^ -0.0015 0.000 (-0.00,-0.00) 0 -0.0014 0.000 (-0.00,-0.00) 0
HlthCard*Q~ -0.0114 0.005 (-0.02,-0.00) 0 -0.0113 0.005 (-0.02,-0.00) 0
(SIEFA_Econ*Q)/1000 0.0173 0.006 (0.01, 0.03) 5.4 0.0198 0.006 (0.01, 0.03) 5.4
  1. ^Dollar values expressed in AUD100,000s.
  2. Reference category is 'working-age adults'. First, second and fourth dummies denote 'young children', 'young adults' and 'older-age retirees', respectively. Here, ∂ UB - UA/∂ xj is for discrete change from reference category to age-group denoted by relevant dummy variable.
  3. Effect(B – A) gives the incremental effectiveness of profile B compared to profile A defined in terms of terms of lives saved for the 'lives-saved' model and life-years saved for the 'life-years saved' model.
  4. ~ For dichotomous variables, ∂ UB - UA/∂ xj is for discrete change in dummy variable from 0 to 1.