Skip to main content

Table 3 Odds ratios (OR) to be specified in the missing data mechanisms given in Table 2

From: Comparison of techniques for handling missing covariate data within prognostic modelling studies: a simulation study

Mechanism

OR

Missingness (%)

for:

for:

5

10

25

50

75

X 2

M X3

101.12

45.5

15.68

5.50

2.17

X 5

M X2

42.04

20.78

7.41

3.00

1.51

X 8

M X3

45.14

14.23

5.44

1.92

0.92

  1. Note: MXi is the missingness indicator for covariate Xi, which equals 1 if an observation is missing and 0 if the value is observed.
  2. OR represents the odds of having two variables with missing observations, and was calculated using the proportion of missing values for each variable and the degree of overlap between variables for each of the five overall amounts of missingness to be imposed.