Another useful diagnostic that gives an indication of the stability of the estimates resulting from multiple imputation is the degrees of freedom (df) associated with the t-value in Rubin's rules and adapted from Schafer [ 8, 9].
However, a common tendency can be found between the quality of the imputation method and the gene cluster stability.
Hence, common trends can be found between the quality of the imputation method and the gene cluster stability.
We applied multiple imputation to investigate the stability of the results due to missing data.
To take attrition at follow-up into account, we verified the stability of the results with sensitivity analyses that used multiple imputation [ 18].
The next step in multiple imputation is specification of the imputation model.
For k-NN imputations, selection of k has an influence on the accuracy of the imputation.
A further improvement of the imputation approach might be to include the unreliability of the WLE as a possible source of uncertainty into the imputation model.
Performance of the imputation process was evaluated using multiple imputation as outlined below [32].
