Figure 1From: Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratioComparison of the relative differences between the 95% confidence intervals obtained by unadjusted Poisson/Cox regression, Poisson regression with scale factor adjusted by χ2 and deviance, Poisson/Cox regression with robust variances and log-binomial regression and the Cornfield 95% confidence interval for each of the six situations studied. S1a (outcome prevalence / confounding): 4.1% / 14%; S1B: 4.9% / 18%; S2a: 31.2% / 8%; S2b: 34% / 17%; S3a: 51% / 4%; S3b: 54% / 25%.Back to article page