Underpowered experiments have three problems: true effects are harder to detect, the true effects that are detected tend to have inflated effect sizes and as power decreases so does the probability that a statistically significant result represents a true effect. To increase power, many researchers only consider increasing the sample size (N). Indeed, standard power and sample size calculations in textbooks and review articles suggest that the only option to increase power is to increase the number of samples. In this article, Stanley Lazic discusses simple options to increase power – often dramatically – while keeping the sample size fixed. The author shows how the design of an experiment and some analytical decisions can have a surprisingly large effect on power.
In the interest of minimising animal usage and reducing waste in biomedical research, scientists should aim to maximise power by designing confirmatory experiments around key questions, use focused hypothesis tests, and avoid dichotomising and nesting, which ultimately reduce power and provide no other benefits.