Skip to main content
Figure 1 | BMC Medical Research Methodology

Figure 1

From: Bias, precision and statistical power of analysis of covariance in the analysis of randomized trials with baseline imbalance: a simulation study

Figure 1

Directional bias of statistical methods. Estimates are given at differing levels of baseline-outcome correlation, treatment effect sizes, and baseline imbalance (−1.96, −1.64, −1.28, 0, 1.28, 1.64, 1.96). Estimates derived from ANCOVA represent the unbiased treatment effect. ANOVA – solid line; ANCOVA – dotted line; CSA – dashed line.

Back to article page