The Art of Statistics: Learning from Data

by David Spiegelhalter

Designing a study that is too small and so has low power, which means that fewer true alternative hypotheses are detected.