Figure 2

( P ₀ and Bias) The bias of the estimate,$p_0^{*}=li{m}_{n\to \infty }{\displaystyle {\sum }_{i=1}^{n}1({\hat{G}}^i=0)/n}$ , depends on p ₀ , μ ₁ , μ ₂ , ${\sigma }_0^2$ and ${\sigma }_1^2$. After fixing, μ₀ = 0, ${\sigma }_0^2$ = 1, and μ₁ = 1, we plot the bias, $p_0^{*}$ - p₀ (y-axis), against ${\sigma }_1^2$ (x-axis), for different values of p₀.