Skip to main content
. Author manuscript; available in PMC: 2014 Aug 5.
Published in final edited form as: Epidemiology. 2010 Nov;21(6):863–871. doi: 10.1097/EDE.0b013e3181f57571

TABLE 3.

Model Goodness-of-fit Statistics by Current Marital Status, Using the Doubly Robust Estimator as the Standard

Prevalence
% (SD)a
Test
Statistic
P
All subjects
  Doubly robust vs. 1.1 (0.1)
    Inverse-probability weighting 1.1 (0.1) 0.85 0.36
    Outcome regression 1.1 (0.05) 0.10 0.75
Married subjects
  Doubly robust vs. 0.4 (0.03)
    Inverse-probability weighting 0.4 (0.03) 0.001 0.97
    Outcome regression 0.4 (0.03) 0.002 0.96
Unmarried subjects
  Doubly robust vs. 12.1 (0.7)
    Inverse-probability weighting 12.4 (0.8) 0.23 0.63
    Outcome regression 12.3 (0.5) 0.19 0.66
a

Prevalence and standard deviation expressed as a percentage.