The biggest differences were observed with RCT 9, which had a large amount of missing data.
Imputing missing data are a necessary step, particularly for the genotypic data sets with a large proportion of missing data per marker (up to 80% in our case).
A large proportion of missing data were present for the ethnic categories.
For studies with a large proportion of missing data, excluding participants with missing data may also reduce the precision of the analysis substantially.
For the six items used to measure Beliefs, the percent of missing data ranged from 11.7 to 12.7%%.
Although the percent of missing data in the former was relatively high, we decided that the bias introduced by restricting the analysis to those observations with complete data would be larger than that caused by imputing missing data combined with appropriate uncertainty analysis [ 62].
The number of completed items, the percent of missing data in every item and the frequency distribution of individual items were determined.
The number of alleles per marker and percent of missing data were calculated for the remaining 380 genotypes with 34 markers (Table 2, Additional file 4: Table S4).
A relatively large amount of missing data on the indicators of pubertal development.
As noted earlier, the case study contained a very large amount of missing data.
