An example

We have data from Bulinus truncatus, a freswater snail to illustrate this multilevel partitioning. When you make the estimates in GDA, however, you may be a little confused because what you'll get is a table that reports $f$, $F$, $\theta_S$, and $\theta_P$. From what we've already seen, you can probably guess that $f$ is our estimate of $F_{IS}$ and the $F$ is our estimate of $F_{IT}$. That means that $\theta_s$ and $\theta_P$ are related to $F_{SR}$ and $F_{RT}$, but how? Well, $F_{RT}$ corresponds to $\theta_P$,¹ and $F_{SR}$ corresponds to

$$\frac{\theta_S - \theta_P}{1 - \theta_P} \quad .$$

So when we run GDA on the Bulinus data we get the results in Table 2. Translating those to $F_{IT}$, $F_{IS}$, $F_{SR}$, and $F_{RT}$ we get the results in Table .

Table 1: Results from a GDA analysis of data from Bulinus truncatus, a freshwater snail.

Parameter	Value
$f$	0.83
$F$	0.87
$\theta_S$	0.24
$\theta_P$	0.04

Table 2: Results from a GDA analysis of data from Bulinus truncatus, a freshwater snail. Translated to equivalent $F$-statistics.

Parameter	Value
$F_{IS}$	0.83
$F_{IT}$	0.87
$F_{SR}$	0.21
$F_{RT}$	0.04