search

Simon-Initiative / oli-torus

Next Generation OLI Authoring and Delivery Platform
https://proton.oli.cmu.edu
MIT License
94 stars 36 forks source link

Sections in Pools Not Working #3768

Closed sandyraysor closed 1 year ago

sandyraysor commented 1 year ago

Module Assessment referenced in video: https://stellarator.oli.cmu.edu/sections/real_chemistry_review_june_24/page/5s48r_compounds_and_chemical_formula_je2lk

Unit Assessment referenced in video: https://stellarator.oli.cmu.edu/sections/real_chemistry_review_june_24/page/5s48r_atoms_molecules_and_ions_z21h1

See video here: https://drive.google.com/drive/u/0/folders/1s_SUGnKNRkqehwuAMkNTwzq89acUOX17

andersweinstein commented 1 year ago

Some notes: in legacy, one can have a pool of sections, each of which contains multiple questions, and a selection with scope = "section", which selects one multi-question section from such a pool for inclusion in an assessment.

Section selection also behaves specially in legacy's targeted retake mode: for retake purposes, a section selected from a pool is only treated as correct if all questions within it are correct. In that way, it behaves like single compound question for retake purposes.

It looks like migration never attempted to handle section selection, which is not directly supported by the torus activity bank model. It also does not give any warning message about an unsupported feature.

The main need addressed by section selection appears to be to define a multi-part question with heterogenous parts, e.g. a numerical question with a multiple choice followup about the same reaction. Section selection and retake behavior allows the author to "fake" such a composite question.

eliknebel commented 1 year ago

We might add support for activity bank groups in the future, but for chemistry this is a content related issue that should be fixed in the XML