Table 2 Sensitivity and specificity of detecting 1.0 standard deviation max(mean differences) across arms (1:3:3 allocation)

From: Choosing an imbalance metric for covariate-constrained randomization in multiple-arm cluster-randomized trials

Imbalance criterion	Resampled				Hypothetical
	max(mean diff) < 1.0		max(mean diff) > 1.0		max(mean diff) < 1.0		max(mean diff) > 1.0
	N	%	N	%	N	%	N	%
KW	44,340	54.15	16,168	89.27	43,800	53.80	18,541	99.73
Inadequate (P <0.30)	44,340	54.15	16,168	89.27	43,800	53.80	18,541	99.73
Adequate (P >0.30)	37,549	45.85	1943	10.73	37,609	46.20	50	0.27
ANOVA	46,555	56.85	15,136	83.57	45,758	56.21	15,356	82.60
Inadequate (P <0.30)	46,555	56.85	15,136	83.57	45,758	56.21	15,356	82.60
Adequate (P >0.30)	35,334	43.15	2975	16.43	35,651	43.79	3235	17.40
MANOVA	18,933	23.12	11,064	61.09	18,630	22.88	11,194	60.21
Inadequate (P <0.30)	18,933	23.12	11,064	61.09	18,630	22.88	11,194	60.21
Adequate (P >0.30)	62,956	76.88	7047	38.91	62,779	77.12	7397	39.79
t test	72,112	88.06	18,111	100.00	69,281	85.10	18,591	100.00
Inadequate (P <0.30)	72,112	88.06	18,111	100.00	69,281	85.10	18,591	100.00
Adequate (P >0.30)	9777	11.94	0	0	12,128	14.90	0	0
WRS	70,287	85.83	17,742	97.96	68,523	84.17	18,591	100.00
Inadequate (P <0.30)	70,287	85.83	17,742	97.96	68,523	84.17	18,591	100.00
Adequate (P >0.30)	11,602	14.17	369	2.04	12,886	15.83	0	0

Abbreviations: ANOVA analysis of variance, KW Kruskal–Wallis, MANOVA multivariate analysis of variance, WRS Wilcoxon rank-sum

Back to article page

ISSN: 1745-6215

Contact us

Submission enquiries: Access here and click Contact Us
General enquiries: info@biomedcentral.com