Skip to main content
Fig. 2 | BMC Research Notes

Fig. 2

From: Genetic model misspecification in genetic association studies

Fig. 2

A Plots for a binary outcome and B plots for a continuous outcome. The sample size required to achieve 80% is lower when the true model is specified (plot 1). The power achieved is higher when the true model is specified (plot 2). And, the shrinkage of the odds-ratio is relatively smaller when the true model is specified (plot 3)

Back to article page