Marks mismatch

Marks mismatch

by Aadil Shah -
Number of replies: 3


We are using moodle 3.9 with coderunner. The problem we face is a mismatch in total marks and the total number of correct questions.

The following is the quiz setup:

There are 40 questions in the quiz each carrying one mark and all 40 questions are python coderunner questions. There is 0 penalty on multiple tries(code submissions).

The above given screenshot of marks mismatch scenario is descried below :

1) Student submits question code and the code passes all tests thereby earns a mark and system records the response.

2) For whatever reason, student submits another version of the code for the same question, for which he /she had already earned a mark, and the code fails the test, the system records the response as wrong answer. Logical until now.


But when the result is generated or the attempt is reviewed this marks mismatch as shown in the screenshot simply produces confusion. 


I tried to change different settings of coderunner and quiz but failed to resolve this. This happens only with coderunner questions.

I may be doing a mistake somewhere, please help me in resolving this issue.


regards

In reply to Aadil Shah

Re: Marks mismatch

by Richard Lobb -
The mark that a student gets for a question is the maximum mark across all their submissions. All Moodle question types behave like this when running with Adaptive behaviour. CodeRunner questions use Adaptive behaviour regardless of the quiz mode settings, in order to present students with a check button to allow them to fix wrong answers.

If a student changes their answer from a correct one to a wrong one and clicks Check, the feedback will say something like:

Marks for this submission: 0.00/1.00. Accounting for previous tries, this gives 1.00/1.00.

The mark shown up by the question number is also the maximum mark over all submissions.

Your screenshot thus shows the the current state of all the question submissions while the total mark shown is the sum of the (maximal) per-question marks.

While I agree this can be confusing, I don't see an easy fix, except perhaps to the wording. I wouldn't want to give a student zero for a question that they had got right at some point but had subsequently altered to a wrong answer. Such changes are almost always inadvertent (why would a student deliberately lose marks like this?) and can occur unexpectedly from such things as typing while the cursor is located in a previous question answer box or while scrolling quickly with the mouse wheel in a Linux system and inadvertently clicking the centre mouse button, which does a paste at wherever the mouse happens to be. Imagine the howls of outrage from students who discovered, when their timer ran out at the end of an exam, that they were getting zero for questions they had got right and didn't know had been changed.

One could perhaps prevent a student changing their answer once they had reached the maximum achievable mark with a given penalty regime. But we sometimes want to encourage students to experiment further even after they've got full marks. We also sometimes do manual style marking and allow students to polish the answer they've submitted.
In reply to Richard Lobb

Re: Marks mismatch

by Aadil Shah -
Thank You for such a great explanation.

I would like to go for the option that you have mentioned in the last para of your reply. Simply prevent students changing their answer once they had reached maximum achievable mark. Please guide me how can i achieve this. Thanks
In reply to Aadil Shah

Re: Marks mismatch

by Richard Lobb -
Sorry, but that's not currently an option. It was a hypothetical solution only, and it's not one I favour implementing at present.