I’m analyzing responses to a pretty long questionnaire in Coda. I’m putting questions in columns, responses in rows. There are few things I wanna figure out if it is possible to do before i go deep.
I want to have metadata about the questions. If the questions were rows, I’d just add a column to provide additional data about those questions, perhaps with a select list linking to another table. But since the questions are columns, I don’t know if this is possible. For instance, I want to categorize the questions, and I also want to provide a short-form version of the question for quick reference along with the full question asked in the row. But if I just add rows at the top providing these additional descriptors then they’ll be treated as answers.
I’m guess it’s not possible to have column headings themselves be look ups to another table, which would solve the problem.
In your case you shouldn’t put answers as columns. Reorganize your table so that each row = one answer to one question. One form would then be a collection of rows with a common value for a “Response” column or so. Then you’ll be able to have your metadata columns that will be per each question.
Something somewhat relevant (see the Replies denormalized table here):
Re column headings — no, but you can have top groups (as you can also see in the linked sample)
Hey Paul, I thought about fully transposing the table before import, but I have a hard time wrapping my head around how that data would be interpreted by the table. Is there another way to explain this? I can’t quite conceive of how I’d have one row per response per question?