Glossary
False discovery rate
The expected share of false positives among the results declared significant. Controlling it lets a researcher screen many ideas with a known share of noise among the survivors.
Benjamini and Hochberg (1995) gave a procedure: sort m p-values, find the largest rank k with p(k) <= k x q / m, and call the first k results discoveries. For independent tests the false discovery rate is then at most q.
It is less strict than controlling the chance of any false positive, so it finds more, and it accepts that some discoveries are false.
Example: with ten p-values and q = 5%, the sorted values 0.001, 0.008 and 0.012 pass their lines (0.005, 0.010, 0.015) but 0.021 does not (line 0.020), so three results are called discoveries.