add clarification about _conditional_ posterior predictive

Author: drbenvincent

examples/generalized_linear_models/GLM-simpsons-paradox.ipynb

@@ -810,7 +810,7 @@
    "source": [
     "The plot on the left shows the data and the posterior of the **conditional mean**. For a given $x$, we get a posterior distribution of the model (i.e. of $\\mu$).\n",
     "\n",
-    "The plot in the middle shows the **posterior predictive distribution**, which gives a statement about the data we expect to see. Intuitively, this can be understood as not only incorporating what we know of the model (left plot) but also what we know about the distribution of error.\n",
+    "The plot in the middle shows the conditional **posterior predictive distribution**, which gives a statement about the data we expect to see. Intuitively, this can be understood as not only incorporating what we know of the model (left plot) but also what we know about the distribution of error.\n",
     "\n",
     "The plot on the right shows our posterior beliefs in **parameter space**."
    ]

examples/generalized_linear_models/GLM-simpsons-paradox.myst.md

@@ -245,7 +245,7 @@ plot(idata1)
 
 The plot on the left shows the data and the posterior of the **conditional mean**. For a given $x$, we get a posterior distribution of the model (i.e. of $\mu$).
 
-The plot in the middle shows the **posterior predictive distribution**, which gives a statement about the data we expect to see. Intuitively, this can be understood as not only incorporating what we know of the model (left plot) but also what we know about the distribution of error.
+The plot in the middle shows the conditional **posterior predictive distribution**, which gives a statement about the data we expect to see. Intuitively, this can be understood as not only incorporating what we know of the model (left plot) but also what we know about the distribution of error.
 
 The plot on the right shows our posterior beliefs in **parameter space**.