Skip to main content


Fig. 5 | BMC Medical Research Methodology

Fig. 5

From: The impact of varying cluster size in cross-sectional stepped-wedge cluster randomised trials

Fig. 5

Illustrative example of the distribution of efficiency in a cross-sectional SW-CRT with 12 clusters and 4 sequences (CVā€‰=ā€‰1.5). CMC: Cluster mean correlation. Efficiency is calculated as the ratio of the precision in a SW-CRT with unequal cluster sizes compared to the precision in a SW-CRT with equal cluster sizes from 44,000 simulations (4000 simulations for each value of the cluster mean correlation). The solid line indicates the median value, the dashed line indicates the lower and upper quartiles, and the dotted line indicates the minimum and maximum values. The black solid horizontal line indicates the reference line. Above the line favours an unequal cluster size design, below the line favours an equal cluster size design

Back to article page