Table 2.

Minimum detectable effect size estimates and summary of parameters used in sample size calculations for the main study outcomes^a

	Women’s dietary MPA	Children’s dietary MPA (3–5 years)	Children DD (6–24 months) (n. of food groups)	Household poultry production	Household poultry sales
Communes (n per intervention arm)	15	15	30	30	30
Villages (n per intervention arm)	30	30	60	60	60
Individuals per village (n)	12	12	8.8	15	15
Intra-cluster correlation (level 3)	0.002	0.002	0.001	0.001	0.001
Intra-cluster correlation (level 2)	0.05	0.04	0.05	0.06	0.01
Sample size (per intervention arm)	360	360	526	900	900
Estimated baseline levels	0.26	0.43	3.95	11	1
Standard deviation (baseline)	0.22	0.31	0.97	11	3
Minimum detectable effect size (SD)	0.28	0.27	0.21	0.18	0.15
Minimum detectable effect size (eq. levels)	0.06	0.08	0.20	1.98	0.45

^aEstimates based on type I error of 5% (two-sided test) and statistical power of 80%

MPA mean probability of adequacy, DD dietary diversity