Formal criteria for the quality of a Theory? Of an evaluation Mechanism?
Formal quality of a Theory?
The quality criteria for knowledge maps in IPA (Wallis 2015) are just
- number of nodes
- amount of connectivity?
So maximum connectivity is when every node is connected to everything else? The best score would just be for zillions of nodes, all interconnected? What would Occam have to say? Surely we know that the lack of connections is an important part of structure (though, as Pearl points out, difficult to code in a correlational approach)?
What formal measures you could use which would give any indication of the actual usefulness of a model
- no orphans.
- “Just enough structure and no more?”
Formal quality of an evaluation Mechanism?
See reliability & validity, bias & illusion.
Wallis, Steven E. 2015. “The Science of Conceptual Systems: A Progress Report.” Foundations of Science. Springer Netherlands. doi:10.1007/s10699-015-9425-z.