Figure 2:

Validity $(\mathit{\text{va}}{r}_m\left(\widehat{\theta }\right)/\mathit{\text{va}}{r}_s\left(\widehat{\theta }\right))$ of Root 1 for both cases, for a variety of classic designs. Ratios above 1 (indicated by blue outlines) correspond to conservative $\mathit{\text{va}}{r}_m\left(\widehat{\theta }\right)$ estimates. Ratios below 1 (indicated by red outlines) correspond to anti-conservative $\mathit{\text{va}}{r}_m\left(\widehat{\theta }\right)$ estimates. For each case, ${\sigma }_t^2=5$ and ${\tau }_t^2=0.1$. To keep the ACC consistent, we chose ${\eta }_t^2=0.1$ and ${\gamma }_t^2=0.05$ for the time-fitted random treatment and treatment-fitted random time cases, respectively.