updated lecture 7 slides

avehtari · Oct 22, 2023 · 379c801 · 379c801
1 parent 86e67ab
commit 379c801
Show file tree

Hide file tree

Showing 29 changed files with 592 additions and 190 deletions.
diff --git a/BDA3_notes.Rmd b/BDA3_notes.Rmd
@@ -712,10 +712,9 @@ $$
 
 - 5.1 Lead-in to hierarchical models
 - 5.2 Exchangeability (a useful theoretical concept)
-- 5.3 Bayesian analysis of hierarchical models
-- 5.4 Hierarchical normal model
-- 5.5 Example: parallel experiments in eight schools (uses
-    hierarchical normal model, skip the details of computation)
+- 5.3 Bayesian analysis of hierarchical models (discusses factorized computation which can be skipped)
+- 5.4 Hierarchical normal model (discusses factorized computation which can be skipped)
+- 5.5 Example: parallel experiments in eight schools (useful dicussion, skip the details of computation)
 - 5.6 Meta-analysis (can be skipped in this course)
 - 5.7 Weakly informative priors for hierarchical variance parameters
     (more recent discussion can be found in [Prior Choice Recommendation Wiki](https://github.com/stan-dev/stan/wiki/Prior-Choice-Recommendations))
@@ -917,7 +916,7 @@ be able to well define the exact tail shape of the prior, but
 half-normal produces usually more sensible prior predictive
 distributions and is thus better justified. Half-normal leads also
 usually to easier inference.
-
+ 
 See the [Prior Choice Wiki](https://github.com/stan-dev/stan/wiki/Prior-Choice-Recommendations)
 for more recent general discussion and model specific recommendations.
 

diff --git a/slides/BDA_lecture_7.pdf b/slides/BDA_lecture_7.pdf