[ARCHIVED] Weighed Grades and Drop Lowest

[corellie]

I have read some Canvas threads on “Weighted Grades” and am still confused.  I have an an assignment group, Exams, that is set to a weight of 50% and has a rule “Drop lowest 1.”  As an example, assume there were 3 exams. Each exam has a different ‘total points.  A student got 10/15 pts on Exam 1, 15/25 on Exam 2, and 5/10 on Exam 3.  What exactly is the algorithm for Canvas to (1) determine which grade is dropped and (2) calculate the students weighted grade for Exams?

References:

https://canvas.ewu.edu/courses/805988/pages/normalize-weighted-assignment-groups

https://community.canvaslms.com/docs/DOC-9880-4152232976

[James · Accepted solution]

 @corellie  

The grade that hurts the student the most is dropped. Not necessarily the grade that has the lowest percentage.

The way it does this is by compute the grade with each of the missing exams and see which comes out best. It keeps that scenario.

1. Exam 1 dropped, use Exam 2 and Exam 3: 15/25 and 5/10 gives 20/35 = 57.14%
2. Exam 2 dropped, use Exam 1 and Exam 3: 10/15 and 5/10 gives 15/25 = 60%
3. Exam 3 dropped, use Exam 1 and Exam 2: 10/15 and 15/25 gives 25/40 = 62.5%

Since 62.5% is the highest, it would drop exam 3.

Technically, there's a complex mathematical algorithm behind that, but it's easier to explain it the way I did.

The grade for exams would be 62.5% and that would count as 50% of the overall grade (assuming your categories add up to 100% -- which they may or may not). The student would get 31.25 points (towards 100%) from the exams.

Note: Edited to eliminate non-standard usage of addition operator.

[Shar]

Thanks James!

I knew you had explained this calculation once before (years ago?) but I couldn't remember where I saw the explanation and my weak search was unsuccessful. I did find the old conversation about keep highest versus drop lowest and that brought back memories, heh.

Cheers - Shar

[kblack]

Indeed! Here's a friendly suggestion to the  @InstContentTeam ‌ to consider another couple of examples to place on that documentation page referenced by the original questioner.

[James]

ishar-uw, 

I've written a bunch of those, so I didn't take the time to bother searching for it. Here are some of the explanations, I'm not sure which one you were thinking of.

• https://community.canvaslms.com/ideas/9718-include-a-keep-highest-scores-rule-instead-of-ignore-lowe... -- this is a long thread, here's the direct link to the post I'm thinking about from December 1, 2018.
• Dropping the lowest grade results in worse score for student 
• https://community.canvaslms.com/thread/23492-dropping-low-score-pulling-grade-down 
• Dropping low assignment score  

There's another one I'm thinking of but can't find.

The other part was about how the 50% factored into and I made a veiled reference to not having 100% in the gradebook. People trying to do extra credit can often throw those calculations off; here are a couple of links to places where I've explained that.

• How do I calculate extra credit or optional assignments in my Gradebook? 
• Unweighted Extra Credit in Weighted Grading Scheme 

[corellie]

Thank you.

I think you meant to say:  Exam 1 dropped, use Exam 2 and Exam 3: (15+5)/(25+10) = 20/35 = 57.14%

Given the algorithm you describe, are all exams of equal weight?  Mathematically, does it penalize a student who got 0/100 more than 0/20 (points gotten / points available)?

I have the following constraint:  "All exams are equal weight."  I scale all exams to 100 (points gotten / points available * 100).  Given the scaled score, I decide with exam to drop. 

[James]

I wrote it in a way that I felt more people would understand, but yes, you add the numerators and add the denominators separately.

The points possible is taken into consideration.

Your statement does not align with the way that Canvas does things. Canvas does not weight them equally unless you weight them equally based on the number of points.

If you want to drop them based off of percentage in your way, then one way to accomplish that is to add another assignment group worth the 50% and make what you have worth 0%. In that assignment group, create assignments that are scaled to 100% and apply your rule to that group.  Another way is to just put the single adjusted grade (after you have manually dropped the lowest) into that category.

Or, you could give the students a boost and let them get the best they could have and not mess with the extra work.

[floresr]

What would happen if a student got two exams with the same low score? would both be dropped?

[DaGIFT222]

Great explanation James!

[DaGIFT222]

Great question!