One potential explanation for fastPHASE's insensitivity to reference sample sizes maybe its fixed small number of clusters.
This again may be due to our limited sample size or maybe due to misclassification bias introduced by self-reported smoking history.
If the sample size is big enough, maybe gender difference will become statistical significant for more end points.
Maybe the sample size is too small to demonstrate the point convincingly?
Although the total sample size of large-scale assessments maybe large, the actual sample size for DIF analysis is less than the total sample size because DIF analysis is an item-by-item approach.
We'd adjust the sample size after the fact, perhaps because some of the mice were outliers or maybe they were handled incorrectly, etc.
Maybe an increased sample size can change this case.
Penta D met Hardy-Weinberg expectations, whereas Penta E did not, maybe because of the small sample size in relation to the large number of alleles of this locus.
Maybe future studies with larger sample size are able to further test this difference.
Unfortunately the low sample size of our study and poor response of rural females maybe, confounding variables.
