False discovery rate (FDR) vs Family wise error rate

The goal of multiple comparisons corrections is to reduce the number of false positives. 
 
False discovery rate (FDR)
- controlling for the proportion of discoveries (significant results) that are false positives
- At FDR of 10%, you're willing to accept 10% of the significant results to be false positives
- use Benjamini-Hochberg

Family wise error rate (FWER)
- more stringent
- At 0.05 family wise error rate, for 100 statistical tests, you will need P < 0.0005 (=0.05/100)
- use Bonferonni


http://udel.edu/~mcdonald/statmultcomp.html