There are several strategies to gain statistical power by manipulating the sensitivity of a study. One way is to attempt to increase the size of the measured effect using techniques such as a targeted user group, distractor tasks, targeted tasks and demonic interventions. However, care must be taken not to mislead with these techniques as they typically have low generalisability. See the noise–effect–number triangle for other ways to increase statistical power.
Used on pages 110, 113, 115, 116, 118