Cluster Design Effect Explorer
Explore the equal-cluster design effect and approximate information-equivalent sample size.
Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.
Explore the simple equal-cluster design effect and the information-equivalent sample size for a specified ICC. All default values are fictional.
Calculation results
Design effect: 1.76
Approximate effective sample: 113.636; actual clusters: 10.
Assumptions and limits
Design effect = 1 + (m−1)ICC. Approximate effective sample = total/design effect. Total must divide exactly into equal clusters. Restricted to nonnegative ICC from 0 to 1. This approximation does not calculate power, an enrollment target, degrees of freedom, or an adequate number of clusters. Unequal cluster sizes, few clusters, varying correlations, repeated measures and complex designs need suitable methods. Fractional effective sample is an information summary, not a count to recruit.
Cochrane: cluster-randomized designs · Read the practical guide · Download blank worksheet (TXT)