Ed-Fi currently has 2 abstractions that capture the concept of academic standards: LearningStandards and LearningObjectives. The concepts are most often used to reflect two different valences:
- Local vs global: LearningObjectives as a localized abstraction (e.g. a district standard) and LearningStandards as shared, governed abstractions (e.g. a state standard).
- Grouping: LearningObjectives as a grouping mechanism for LearningStandards.
In field work, these valences are often combined, and hence the abstraction. However, this abstraction creates several issues:
- it complicates the model: it makes the model - and a recursive part of the model - a lot more complex in terms of possible relationships
- it introduces ambiguity into the model for the same reason, as both mappings are possible. Which does someone use, and what are the consequences if someone else does that differently?
In terms of other ecosystem specs, it is worth noting that both CASE and Academic Benchmarks seems to have a model with a single entity type. In this case, a mapping from CASE or AB to Ed-Fi would have to make what seems like some pretty arbitrary choices about when the model went from LearningObjective to LearningStandard.
In terms of alternatives to capturing these valences listed above:
- The concept of "local vs global" might be reflected in a namespace that clarified the origin of the concept/skill being measured.
- The model will naturally reflect through relationships the notion of grouping, and no further signal of this should be necessary