Is there a way to change the survey languages for participants more economically than one by one in a participant table of an an active survey? I would need to change hundreds of languages from default to another.

You can export as a file, editing the language with an editor and then import the modified file.

Would this work with additional attributes in an active survey as well, and not result in a loss of response data?

You can export the token table, edit the language column in something like LibreOffice Calc and then re-import the table (over-writing the existing table) without affecting the data.

I recently did this for a client who forgot to assign another language for 1/2 of the tokens. We did as described above and also modified the "Invitation sent" column so we could resend the invitations in the appropriate language to those with the new language.

Thanks. So, do I understand right there is two ways to edit token table attributes "on the fly" in active survey. First by updating the token table by sql code davebostockgmail wrote in another string:

As an example .. if the Identifying Variable is Attribute1 in the table and the one you wanted to update was Attribute5 then the sql code would be....
UPDATE `surveys`.`[Your Token Table]` SET `attribute_5`='1' WHERE `attribute_1`='GroupA';
This will update the table for all participants who have a value of GroupA

In my case I can identify participants whose attribute (=survey language) needs to be changed based on value in one of the additional attribute.

And second method would be exporting the whole token table and re-import an edited version of it.

So should I delete the existing token table before re-import of edited csv.file? Or just import?

You could also use my MassAction plugin, which gives a simple Excel-like edit mode for questions, question groups and participants/tokens.

github.com/olleharstedt/MassAction

@Trifon, I think you should see an option to over-write (replace) the existing tokens when importing. I'm not in a position to test this week so you should try it with a dummy survey first.

@olle, nice! I'll test when I get back next week.