E.g. survey 188183011, page 3, question 1. I can try addressing the same way we treat question_type == "multiple choice", we can add the condition question_subtype == "multi".
I'm less clear what should go in the answer cell. Say you it's a survey about which foods you eat for which meals and you ultimately have the combined column header "dinner - cereal" created from a multiple choice matrix of column = meal and row = food. Should the cell value be cereal? Dinner? Right now it's the column, so dinner.
I'll implement that and leave this issue open. Related issues:
Is it friendlier to return a list-column for multiple choice questions? I think it's more elegant but less manageable for the lay-user.
Why is the function for parsing multiple-choice columns into a single table, tntpr::check_all_recode(), performing weirdly here?
E.g. survey 188183011, page 3, question 1. I can try addressing the same way we treat question_type == "multiple choice", we can add the condition question_subtype == "multi".
I'm less clear what should go in the answer cell. Say you it's a survey about which foods you eat for which meals and you ultimately have the combined column header "dinner - cereal" created from a multiple choice matrix of column = meal and row = food. Should the cell value be cereal? Dinner? Right now it's the column, so dinner.
I'll implement that and leave this issue open. Related issues:
tntpr::check_all_recode(), performing weirdly here?