Sample Size and Statistical Power

Even if we generate an appropriate sam­pling strategy, we need to consider the sample size that is necessary for the proposed analy­ses. Sample sizes vary greatly across studies: In some of the studies reviewed for this chapter, sample sizes were inadequate for the methods used; for example, some studies attempted to estimate many parameters via analysis of variance (ANOVA) or multiple regression on samples with as few as 20 participants. The difficulty in obtaining appropriately sized samples should not be used as an excuse to do analyses that lack statistical power.