Html5 retry quiz (on iPad) does not wipe users score (among other oddities)
Mar 22, 2016
Testing was performed with iPad from SCORM Cloud.
So if I take this exam, or any others I've tested since last week (previously was testing with no problems, though I cannot prove this behavior is new), and I achieve the minimum passing score of 80%, then hit "retry quiz" AND THEN leave the screen(close the quiz/exam window), the system does not act as I would assume it should.
I would think that it should then wipe any previous score (actually would think it should wipe the score when I hit "retry quiz") and communicate a status of "incomplete" back to the lms.
The actual behavior is like this: I complete with an 80%, click "retry quiz" and am presented with the first question of the quiz, absent of answers(though if I look at the "question list", it still shows all of the correct and incorrect icons from the previous score**). At this point:
If I do not answer any questions and just close the window, the package tells the system to register the previously attained 80% along with a "passed" status (though all scores should have wiped when I clicked "retry quiz").
If I do answer any questions before closing the window, then the system will wipe the previous score of 80%, but will then accurately grade only those questions answered on that second attempt. For instance, if I then aswer 1 of 5 correct, leave 4 of 5 blank, and close that window, I am returned to the lms and see that a passing status and score of 20% is now locked in as permanent. Minimum passing is set to 80%.
Behavior seems to work ok** if I actually complete the quiz when I choose to retry, but seems broken when I choose to retry and then leave early.
On desktop browser, functionality seems fine.
**an update to iOS 9.3 has rendered the question list access broken for me as well. If I tap on it, it briefly blinks open and then closes, not allowing me to actually navigate by clicking on a Q in the list. It had been temperamental prior to the update(taking a practiced touch to operate), but I could use it, now, not so much(was hoping the update would fix the original problem, but did not help).
Attached are the proj and zipped package files.
Can you duplicate and/or provide any insight? Thanks!
7 Replies
Hi Patrick!
Most learning management systems (LMSs) consider a course to be in review mode when attempting it again after the completion requirement has been met. This prevents learners from accidentally changing their course status from complete back to incomplete. If you'd like to change this default behavior, please review this article for details.
Thank you for sharing your file. I am updating to 9.3 and will test your file when complete.
Hi Patrick!
I was able to update to 9.3 and test your file.
I'm not seeing the behavior you are reporting:
My test file is here.
Leslie, first off, Thank You! You folks have been, and continue to be a lot of help to us.
Also, this is getting kind of deep into detail, so to try and keep from over-complicating I do plan to test the file you shared, with regard to the question list, but any follow up on that will be moved to a new thread.
Ok, so I don't think I explained myself very well the first time, but the issue is less about me disagreeing with the way the quiz behaves than it is about getting inconsistent results from the same quiz when accessing via different environments (SCORM Cloud via Chrome on pc and via Safari on iPad).
Looking closer, it seems that I am having an issue where the flash version and html5 version of the same QM 13 quiz are behaving differently from one another, though only in certain situations. I went ahead and logged some quick testing I performed on SCORM Cloud to better illustrate this. Please see columns (L) and (M) to see the different results obtained by performing the same behavior.
I do think that you have already put me on the right track to preventing and/or working around this(link above), but in and of itself, this seems like an unintended flaw that ya'll might want to look into further.
As for actually fixing it, I'll have to test more before I know for sure, but I don't believe that changing "var REVIEW_MODE_IS_READ_ONLY = true;" to false is what I need (prelim testing did not show the results I expected/needed, but I will test further and log results to make sure I'm not confusing myself).
That said, I do think the Configuration.js file may have the fix I need. Once I again, I've only done limited testing of this, but I think switching this: "var SCORE_CAN_ONLY_IMPROVE = false;" to true, may indeed be a suitable workaround/fix (at least, it should prevent the system from issuing/communicating a passing status alongside a non passing score - again, please see column (M) for more detail).
Thanks again, I'm attaching the testing spreadsheet I referenced, as well as the project and zipped package files being referenced.
Thank you for such detail Patrick. I will share this with our QA Team for further review.
Hello,
I am very interested in this as well as we have been working on solutions to odd behavior in html5 content; though I do not have the impressive, copious logs that Patrick has created to back up my observations. Thank you for your work in this area Patrick.
By commenting here Steve, you should now be subscribed to subsequent responses in this forum thread. The report filed includes this thread so that we can update when applicable.
:)
This discussion is closed. You can start a new discussion or contact Articulate Support.