It is well known that this correction is too conservative and reduces the power dramatically and unnecessarily [88]. Controlling the false discovery rate rather than the overall false positive rate has been proposed as a less conservative method. The false discovery rate is the proportion of false positive associations among the detected significant associations and can be controlled for using a simple algorithm [89]. Work conducted in the past few years to reduce the number of falsely significant associations in microarray data analysis also provides a variety of solutions even in small samples with correlated data [88,90–92]. Neither procedure changes the rank of the P-values, but simply provides additional guidance as to which associations are most significant across the entire study. Some statistical methods are more powerful than others and often a substantial increase in power can be accomplished by adopting more sophisticated statistical analyses without needing to increase sample size.
