MI involves specifying a parametric model for the missing data given the observed data and drawing missing values from the posterior predictive distribution M>1 times.
Pairwise deletion was used to remove specific missing values from the analysis.
Even though the missing values from the external jugular vein were high (32.5%), we have data from 11 subjects, which is sufficient according to our power calculation.
In the panel survey in 2012, CFPS supplemented this question, allowing us to draw the missing values from the new responses.
After removing the missing values, from an initial sample of 2000 randomly selected individuals between the age group of 18 64 years the final sample used for this study contains 1961 respondents.
There were no missing values from the dataset.
While most SNPs had fairly complete data, others had missing values from substantial (up to 28%) numbers of subjects.
In this way, the word processing times were not distorted by missing values from slow or fast participants.
In these cases we extracted the missing values from the primary data and "filled in" the values to construct the patterns of gene expression across all seven time points.
When missing values occur because the records do not contain the sought-after information, researchers should first attempt to establish the meaning of missing values from external data whenever possible (e.g., from historical weather data, see Uttl et al. [15]).
We excluded missing values from the analyses.
