ddm_hierarchical_tutorial

[mariama-design]

No description provided.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[frankmj]

Looks great! Some suggestions:

 1. Overall can include more comments for educational purposes

* best to include a little preamble at the top explaining the different situations you will model

* make it more concrete, so replace "x" with "difficulty"

* explain to user that the forward simulations (with do operator etc) are for the sake of generating fake data so that we can test if we can recover the true parameters, but if one only has their own data to estimate, they can go straight to the step where they fit it (but that it is always a good idea to check that one can recover parameters from synthetic data).

* when showing the model graphs put a link somewhere to documentation (from elsewhere in HSSM tutorials) on how to interpret these (deterministics etc)

2. For PPCs, it would be better to show how these vary by a key factor in the experiment. So plot a PPC not just for each response, but separately for different levels of x (difficulty) in variant A, and levels of age in variant B, etc.