Sort final data.frame by the same thing that the .xlsx export is sorted on

[sfirke]

So the rows look the same when comparing them. Looks like that's response_id, in descending order.