Next section: Size of difference between means

The factors affecting power will be illustrated in terms of a simple example. Suppose an experimenter were interested in testing whether a drug affects reaction time. Subjects are tested once after taking the drug and once after taking a placebo (naturally, half of the subjects take the drug first and half take the placebo first). The difference in reaction time between the drug and placebo conditions is calculated for each subject. The null hypothesis is that the population mean difference score is zero. H

μ

To simplify the calculations, assume that the standard deviation of the difference scores (σ) is known to the experimenter. The experimenter samples N subjects at random and computes the mean difference score (M

Subsequent sections will use this example to investigate the factors affecting power.

Next section: Size of difference between means