Most QTL detection studies have used either maximum likelihood or regression
methods to test for segregating QTL.
Generally, the statistical analysis is computed for each marker on the basis of the null hypothesis that there are no segregating loci, affecting the trait of interest, linked to the genetic marker.
A null hypothesis is rejected when the probability of the test statistic is below a predetermined level, usually 5% or1%. However, if many markers and traits are tested, several null hypotheses will meet these rejection criteria by chance .
Several studies have suggested that only effects that display statistical significance on two independent trials should be considered as confirmed .