Calculate text response from repeat outside repeat in web-form not working

[KateSherman_psmfc]

I have a survey where users are filling it out in a web-form. I have a repeat where I asked for information about organizations working on a restoration project. I am able to create a field that joins all of the organizations into one list (outside of the repeat). However, I want to have a field outside of the repeat that is calculate if a user select the organization type as the "project sponsor." I am able to get this to work if the user selects the project sponsor/lead first in the repeat. However if they add it 2nd or 3rd, anywhere other than first record in the repeat, it does not work. Is this a bug, any suggestions for a workaround?

Attached the survey for reference. I'm working with org_sponsor field within the repeat and organization_lead outside of the repeat.

[abureaux]

Sorry, I'm not sure I understand the functionality you are looking for.

I understand that this is the repeat in question, and the yellow fields are the fields of interest. Could you please elaborate some more on what the desired behaviour would be?

On a side not, the [hidden] type does not accept any appearance. It's basically just a raw text field. If you want more control over hidden fields, you could specify a type like normal, and then use the [hidden] appearance.

[KateSherman_psmfc]

Yes, those are the fields, however, organization_role also plays a part. Once someone selects "project_sponsor" from the organizational_role field, it should populate the name of the organization in the hidden org_sponsor field. There should only be one project_sponsor, and I want that to be calculated outside of the repeat in the main feature class. 

Basically, there can be many organizations involved in the project (funder, technical support, project sponsor/lead). Each project should have a project lead, and I want to separate that particular response out.

Thanks for the information about the hidden type, I didn't realize it can't support appearance.