Hello Canvas People,
I have a question regarding the way in which Survey Data is exported from Canvas. We offer a satisfaction survey in all of our courses and when it is exported from Canvas as a .csv file, the data has a lot of extra columns, special characters, and random numbers requiring a lot of manual clean up. Does anyone know a way around this and/or why this is occurring. I am using the "survey statistics" "student analysis" feature to generate my reports.
I have attached a picture below for reference. The highlighted numbers appear in front of every question, followed by a column of numbers 0,1 and column AE shows one of the special characters that almost always appears on export.
Any insight or suggestions would be appreciated. Thank you.
Tech Support Coordinator, Continuing Education
firstname.lastname@example.org Office: (321) 674-8382 / Fax: (321) 674-7050
Florida Institute of Technology
150 W. University Blvd. Melbourne, FL 32901
Those aren't random numbers, they're the internal Canvas question ID and the text of the question follows, but it probably has an embedded newline so you can't see it unless you do something that makes the text wrap.
For the weirdness in column AE, it appears to be an essay response question. This means that students can type in whatever they want. Some prepare their comments elsewhere and then copy/paste into the box and it's possible that what they're copying includes some issues. There was another time in another question where someone was providing suggested responses like "Enter N/A if not applicable" and the student's were highlighting the item the instructor had supplied. It had some special coding there, so you might want to take a look at the actual questions.
The 0's and 1's are useless for a survey data that is not scored and doesn't have right or wrong answers and can be ignored.
You might find these items helpful in understanding what's going on and how to deal with it. You're not the only one frustrated with the student analysis report.
As messy as it is, though, I'm not sure of a better way to do it. You're trying to represent a potentially complicated process in a two-dimensional structure. I think better support for the question names (Canvas asks for them but then never really uses them) might be more useful, but they aren't necessarily unique and people may still want to see the text of the question. You're also limited in that the CSV dump is a text-only format and the original questions might have had images in them. Surveys may not have right and wrong answers, but you want to use the same format for all of the student analysis dumps. It goes on and on, and for almost every issues there's a justification for the reason it is the way it is and a problem with the way someone else wants it to be. Unfortunately, to make data consumable and understandable by the masses, it sometimes gets presented in a way that makes a deeper analysis not so easy.