Experiments with multiple nested levels where randomization can take place at any level bring challenges to the computation of sample sizes. Formulas derived under simple single-level experiments must be adjusted using multiplicative factors or design effects. In this work, we take a unified approach to finding the design effects in terms of intracluster correlations and present formulas to compute sample sizes of different levels. Equal cluster sample sizes and homogeneous within cluster variances are assumed.
MaasCJMHoxJJ. Sufficient sample sizes for multilevel modeling. Eur J Res Meth Behav Soc Sci2005; 1: 85–91.
2.
ScherbaunCAFerreterJM. Estimating statistical power and required sample sizes for organizational research using multilevel modeling. Organ Res Meth2009; 12: 347–367.
3.
MurrayDM. Design and analysis of group-randomized trials. New York, NY: Oxford University Press, 1998.
4.
DonnerAKlarN. Design and analysis of cluster randomization trials in health research. London: Arnold, 2000.
5.
HeoMLeonAC. Statistical power and sample size requirement for three level hierarchical cluster randomize trials. Biometric2008; 64: 1256–1262.
TeerenstraSLuBPreisserJS. Sample size considerations for GEE analyses of three-level cluster randomized trials. Biometrics2010; 66: 1230–1237.
8.
CornfieldJ. Randomization by group: a formal analysis. Am J Epidemiol1978; 108: 100–102.
9.
RaudenbushSWLiuX. Statistical power and optimal design for multisite randomized trials. Psychologic Meth2000; 5(2): 99–213.
10.
PituchKAMurphyDLTateRL. Three-level models for indirect effects in school- and class-randomized experiments in education. J Exp Educ2010; 78(1): 60–95.
11.
GlynnRJBrookhartMAStedmanM. Design of cluster-randomized trials of quality improvement interventions aimed at medical care providers. Med Care2007; 45: S38–43.
12.
RufDBernerMKristonL. Cluster-randomized controlled trial of dissemination strategies of an online quality improvement programme for alcohol-related disorders. Alcohol Alcohol2010; 45: 70–78.
13.
GrandesGSanchezASanchez-PinillaRO. Effectiveness of physical activity advice and prescription by physicians in routine primary care. Arch Intern Med2009; 7: 694–701.
14.
KristAHWoolfSHRothemichSF. Randomized trial of an interactive preventive health record to enhance the delivery of recommended care. Ann Fam Med2012; 10: 312–319.
HedgesLHedbergEC. Intraclass correlation values for planning group-randomized trials in education. J Educ Eval Pol Anal2007; 29: 60–87.
17.
ThompsonDMFernaldDHMoldJW. Intraclass correlation coefficients typical of cluster-randomized studies: estimates from the Robert Wood Johnson Prescription for Health Projects. Ann Fam Med2012; 10(3): 235–240.
18.
CookAJBrucknerTMacLennanSG. Clustering in surgical trials – database of intracluster correlations. Trials2012; 13: 2–8.
19.
ShihWJ. Sample size and power calculations for periodontal and other studies with clustered sample using the method of generalized estimating equations. Biometric J1997; 39: 899–908.
20.
DangQMazumdarSHouckPR. Sample size and power calculations based on generalized linear mixed models with correlated binary outcomes. Comput Meth Program Biomed2008; 91: 122–127.
21.
EldridgeSMAshbyDKerryS. Sample size for cluster randomized trials: effect of coefficient of variation of cluster size and analysis method. Int J Epidemiol2006; 35: 1292–1300.
