Downloading all responses in a New Quiz, including all of the file submissions

tgideons
Community Participant

Before I start weeping in bitter frustration, I need to confirm what I am now suspecting.

Question 1: Is it possible to download, in bulk, all of the files submitted by students in a New Quiz?

Background: I just gave a final exam to 260 students, and each exam had two essay questions. Because Canvas refuses to link their essay questions to Turnitin, I have had students upload files, which I then download in bulk and upload in bulk to Turnitin. In the past, this workaround worked pretty seamlessly in the Classic Quizzes, except for the fact that students could upload any file type they wanted to, no matter how many times we tell them to only upload in a particular format. Canvas, despite what appears to be numerous demands over many years to give us the ability to limit the formats in the file submission question, refused to fix this in Classic Quizzes. I found out yesterday that they did it in New Quizzes, so I switched the exam to New Quizzes. Yeah, that was stupid of me.

Now that the exam is over, I have been looking for a way to download the essays. Based on my searching, it appears that this is not possible in New Quizzes, but I cannot determine this for sure. I was only able to find reference to the issue in forum questions and complaints, as there is no documentation from Canvas saying one way or the other. Canvas documentation tends to deliberately leave out key information like this, so I shouldn't be surprised. (And then the tech support reps will tell me it's my fault for not knowing that Canvas doesn't have basic functionality.)

Question 1a: If it is not possible to download file submissions in bulk, has anyone figured out a workaround that will prevent me from manually downloading 520 files one by one over the next week? 

Question 2: Is possible to download the full quiz and item analysis reports as .csv files?

Background: Like many, many other people in this forum, having the ability to access the quiz answers of all of our students in one spreadsheet is absolutely essential to our jobs. It appears that this feature is not available, but again, I have found it impossible to determine for sure because the Canvas documentation says nothing, and I can only triangulate through responses to forum posts and mostly opaque blog posts about improving the reporting functionality through an API. And by nothing, I mean that the page in the documentation of the reports doesn't mention anything at all about how to access the data, let alone download it.

And I really need access to the full quiz data -- including all the responses to all of the questions, as we had in Classic Quizzes -- because I believe it is the only way I will be able to determine which of my students were served the same questions twice. What happened? I built the exam using several blocks of questions. Questions 1 through 8 were randomized from Bank A, Questions 9 and 10 from Bank B, 11 and 12 from Bank C, and Questions 12 through 20 were also from Bank A.

I tried to determine if this construction would cause any problems, but there is no mention of it in the documentation on building the quizzes. I assumed, foolishly, that of course no one would allow the same question to be randomized twice into an exam, but, as I just indicated, I am a fool to think such things. I'm even more of a fool to think that Canvas would serve a warning to a user that this might happen. When I contacted tech support, they knew it happened, knew it was a problem, and thought it might be worth getting into the documentation. Ya think?

Anyway: Several students were served repeats, but I can't determine how big a problem it is without either manually going through all 260 exams one by one in SpeedGrader or, preferablym having access to a simple .csv file of the quiz data that I can analyze in Excel. 

Question 2a: If Canvas has not decided to grant us access to our data, have any users figured out a workaround? 

Thank you. I'm going to go take a benzo.

Users who also had this question