False Discovery Rate (FDR)

Authors

Affiliations

Nate Yomogida, B.S., SPT

Doctor of Physical Therapy

B.S. in Kinesiology

Chloë Kerstein, B.S., SPT

Doctor of Physical Therapy

B.A. in Neuroscience

false discovery rate (FDR) is the expected value of V/R, where V is the number of true hypotheses that are rejected (i.e. the number of false rejections) and R is the total number of rejected hypotheses¹

total R is known and equals V þ S, where S is the number of rejections of false hypotheses, i.e. of correct rejection

Due to this definition, $FDR\le FWER$, since if FWER = a, V/R is zero with probability $1-\alpha$ and between 0 and 1 with probability a, and thus its expected value is $\le \alpha$¹

References

1.

Shaffer JP. Controlling the false discovery rate with constraints: The Newman-Keuls test revisited. Biometrical Journal Biometrische Zeitschrift. 2007;49(1):136-143. doi:10.1002/bimj.200610297