Figure 2From: Is there a role for expectation maximization imputation in addressing missing data in research using WOMAC questionnaire? Comparison to the standard mean approach and a tutorial Box-plots of the 3 WOMAC subscale simulations using the dataset with a random sample of complete cases (N = 200) and 5 times the missing rate. Numbers on the X-axis refer to the median and range of samples considered in the score evaluation for different imputation methods Numbers on the Y-axis represent the range of scores of a particular subscale.Back to article page