Fig. 1From: Negative log-binomial model with optimal robust variance to estimate the prevalence ratio, in cross-sectional population studiesComparison of the standard errors of estimation of the PR adjusted by age groups, between the models with robust variance and MHBack to article page