The use of a single imputation to fill in missing data is unbiased or minimally biased quite generally but suffers from inaccurate estimates of variance and, consequently, CIs that are too narrow, particularly when missing data exceed about 30%.
In all analyses only questionnaire data with complete answers to all items comprising a subscale were analyzed to ensure that analyses were unbiased by missing values.
MCB: missing call bias; UBM: unbiased missing; SNP: single nucleotide polymorphism; QC: quality control; HWE: Hardy-Weinberg Equilibrium.
The overall data set has approximately 20% missing data, but gives unbiased estimates of parameters based on Migrate-N 3.2.6.
However, methods for dealing with missing data to achieve unbiased statistical results, such as multiple imputation or doubly robust estimators, are increasingly widely used, and these rely on models that effectively predict missing data.
International literature proposes many papers and reports dealing with GEN IV NPPs, but there is an evident difference in the types and structures of the information and a general unbiased overview is missing.
The observed genotype frequencies can be presented as p G i o b s = c i p G i ∑ i c i p G i, where the overall call-rate is defined as C a l l R a t e = ∑ i c i p G i. When c AA = c Aa = c aa, there is no bias in missing data, we called it Unbiased Missing (UBM).
The model-based estimates are unbiased with unbalanced and missing data, so long as the missing data are non-informative (missing completely at random).
Complete case analysis excludes participants with missing data and is valid (produces unbiased estimates) if missingness is independent of the outcome, given covariates.
Under the assumption of data missing at random (MAR), FIML provides unbiased parameter estimates compared to traditional methods (such as last observation carried forward) even with a substantial proportion of missing data [ 51].
