A way to increase statistical power. Variability due to factors that are not controlled or not known about is called noise. Noise can be reduced by (1) controlling conditions in a study to limit variability and (2) obtaining information about factors that may affect results and modelling these effects.
Also used in hcistats2e: Chap. 13: pages 149, 153, 154, 156, 162
Used in glossary entries: statistical power