Fig. 1From: Studying missingness in spinal cord injury data: challenges and impact of data imputationSimulation study overview. A The first step leads to the definition of 500 bootstrap subsets, with n = 500 in each subset; B In each bootstrap subset, 30% missing data is introduced in the variable for lower extremity motor score (LEMS) (either at baseline or at recovery) according to one of the three patterns of missingness (missing completely at random [MCAR], missing at random [MAR], missing not at random [MNAR]), independently, before being imputed. Empty circles represent missing entries, while plain circles represent known values; AIS: American spinal injury association impairment scaleBack to article page