[SAK-36785] Multiple choice multiple correct giving 0 points Created: 17-Apr-2017  Updated: 17-Apr-2018  Resolved: 03-May-2017

Status: Verified
Project: Sakai
Component/s: Tests & Quizzes (Samigo)
Affects Version/s: 11.3, 12.0
Fix Version/s: 11.5 [Tentative], 12.0

Type: Bug Priority: Critical
Reporter: Kenneth Aragon Assignee: Kenneth Aragon
Resolution: Fixed Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
is related to SAK-34373 Samigo answer is correct but score is 0 RESOLVED
is related to SAK-36825 Changing option in multiple choice mu... Verified
11 status: Resolved
Previous Issue Keys: SAM-3191


When using the Multiple Choice Multiple Correct question type and having a point value of 1.66 results in the Student getting 0 points even though all questions have been answered correctly. Steps to reproduce:

1. Login as Instructor
2. Add the Tests and Quizzes tool
3. Create a new exam
4. Add a Multiple Choice question type
5. Select the Multiple Choice Multiple Select radio button
5a. Select the 'All or Nothing' option
6. Add a question and possible answers
7. Check the Correct Answer box next to a couple of those answers
8. Publish the Exam
9. Take the exam as the Student and select the correct answers
10. Submit the exam

Expected result:
User should receive a full 1.66 score for that question

Actual result:
User receives 0 points even though all answers are correct.

Note: If I update the point value to be an even 1.0 and regrade / republish, the student then receives credit.

Comment by John Ansorge [ 27-Apr-2017 ]

I've seen something very similar happen. Changing the points a question is worth affects whether the MultChoiceMultCorrect is actually scored as correct even when all answers are correct. That suggests to me that there's some kind of potential for (floating point?) rounding errors in this determination if a answer response is "correct".

For example, if you create a question with 7 answer choices and make 6 "correct" and one "wrong" and set it to all-or-nothing grading, it will be scored accurately if the question is worth 33 points or 15 points. It will always be scored inaccurately if it's worth 1 point or 100 points. You can show this by editing a published exam students have taken and changing the point values of the questions. The exact same responses will go from getting zero points to full points.

Comment by Matthew Jones [ 03-May-2017 ]

Patch merged, assigning back to Kenneth Aragon for verification after it gets into nightly.

Comment by Sam Ottenhoff [ 11-Jul-2017 ]

Can someone verify?

Comment by Gnapika Reddy Kudumula [ 12-Jul-2017 ]

Tested the issue on https://trunk-mysql.nightly.sakaiproject.org/portal/, the issue is fixed. 

Comment by Jeff Pasch [ 17-Dec-2017 ]

Hi Kenneth Aragon and Neal Caidin. This one just bit us in production on 11.2. What do you think about making this a Critical since it affects student scores? Also seems the affects version should be at least 11.2 if not 11.0. Thanks.

Comment by Neal Caidin [ 18-Dec-2017 ]

Hi Jeff Pasch , in one sense it doesn't matter what the priority is set to since this is already fixed and incorporated into 11.x, 12 and 13. But if you want to set it to Critical, I don't have any objection. 


Comment by Matthew Jones [ 18-Dec-2017 ]

Yeah, priority is "generally" just used to try to push developers attention to specific issues. After they're fixed the priority isn't as referenced as often other than possibly blockers. I'm happy going to Critical with this though. I think all schools at this point should be running 11.x and not the last released branch, 11.4 as there are almost 50 blocker/critical issues since then.

Comment by Neal Caidin [ 18-Dec-2017 ]

Matthew Jones , that definitely argues for getting an 11.5 out , perhaps after we release 12.0 ?


Generated at Sun Feb 23 11:41:05 CST 2020 using Jira 8.0.3#800011-sha1:073e8b433c2c0e389c609c14a045ffa7abaca10d.