Statistical power and large sample sizes

Large sample sizes increase statistical power and increase the flexibility of the effect size

Statistical power will always increase with large sample sizes. Large sample sizes allow for both large and small effect sizes, regardless of limited or extensive variance in the outcome, to be detected (flexibility).

Large sample sizes provide enough statistical power to find treatment effects in between-subjects, within-subjects, and multivariate designs, as well as when outcomes are measured at continuous, ordinal, or categorical levels.

​From a data analysis standpoint, there are many things one can do with data. However, remember that experimentwise error rates (Type I errors) increase with more analyses.  
Large sample sizes increase statistical power and increase the flexibility of effect sizes.
Large sample sizes increase statistical power and increase the flexibility of effect sizes.
Click on the Sample Size button to continue.