Instead of a creating a participant record for every participant-transcript instance, create one unique participant based on <corpus_id, code, name, role> fields. This was done as an attempt to make the Participant table more intuitive, as there is now only one instance per person (or at least that's the way it should be). The age is now being recorded in the Transcript, Token, and Utterance tables.
Use the following query to see potential pitfalls with this approach:
select *
from participant p inner join participant p2 on p.name = p2.name and p.corpus_id = p2.corpus_id
where (p.role != p2.role or p.code != p2.code)
group by p.corpus_id
Instead of a creating a participant record for every participant-transcript instance, create one unique participant based on <corpus_id, code, name, role> fields. This was done as an attempt to make the Participant table more intuitive, as there is now only one instance per person (or at least that's the way it should be). The age is now being recorded in the Transcript, Token, and Utterance tables.
Use the following query to see potential pitfalls with this approach: select * from participant p inner join participant p2 on p.name = p2.name and p.corpus_id = p2.corpus_id where (p.role != p2.role or p.code != p2.code) group by p.corpus_id