Random Points & Imprecise Percentages

I finished reading Thomas Guskey’s On Your Mark last night in preparation for a discussion we had today on grading as a part of our PD at Brookfield Central.  As a part of the preparation for the discussion, we were asked to read an article on grading and share out our thoughts in an Ed Camp style environment. Different rooms focused on different practices.  I was shocked to realize we had been talking for almost an hour when our time was up.  I felt like we had just begun digging deep.  It was a great feeling to hear these conversations occurring with a positive mindset.

Every once few months I read, see, or hear something that really makes me realize the box I’ve placed myself in.  The box that has been revealed to me this week is the box of using percentages to guide grading.  Like anyone trying to look outside of the box they are in, these views may seem a bit fragmented.  It’s my hope that putting them down will help me see the logic in the truth Guskey puts forward.  So here are a bunch of random points that are circling in my mind.  Someday soon, I hope to comeback and refine them into something a little more succinct.

Again, I don’t claim any of this insights to be original.  It’s just me understanding the conversations that have been out there for a while now.

• Our current report card reports out grades of A - F.  That’s 5 distinct grade levels with +’s and -’s which don’t affect the true grade point (4 - 0 scale).  Our current grade book translates percentages into those grade categories.  Our percentages go down to the hundredths place.  That’s 10000 distinct scoring possibilities.  

• We have the ability as teachers to decide how many points an individual assignment is worth.  These points are then turned into a percentage by the grading program and then into a grade.  By making an assignment worth more points, we can create more room for variance.  But do each of these points actually alight with a piece of feedback.  If I have a 100 point assignment, do each of these 100 point align to a specific piece of feedback or have I just made it 100 points to show it was worth more than a smaller assignment or is it because I am aligning it to a 100 percent scale.

• If ultimately we are translating our grades to a 5 level scale, A - F, why are we working with a 100% scale?  

• Do I dare mention the cut off for a “D”?   Let’s say the cutoff for a D is 60%, should I be designing all of my assessments so that a student who gets 60% of the total points has demonstrated the bare minimum for passing?  Then, I need to align every grade to each percentage range.

• It makes more sense to me to decide what the grade is outside of the percentage and not add that translational step between the assessment and the grade.  Why should an assessment be run through a percentage filter before getting a grade?  

• Have the students score a distinct grade not points.  Or, if we want points, let’s stick with 4 - 0.  Once they are given that point score, the teacher can provide specific feedback including celebrations and suggestions for improvement.  

• I understand that a percentage may be important for compositing grades across assignments.  

• Guskey does a great job of pointing out that we have this 100% system to give us a sense of precision in our reporting of grades.  But, it provides an illusion of precision.  Does an 84% vs an 85% provide actionable or meaningful feedback for a learner?

• Let’s look at my district’s different grading scale options standard curve which is similar to many traditional percentage scales I have seen.  

Grade	Standard Curve Percentage	Integer 5 Point Scale	Integer 4 Point Scale
A+	100-99
A	98.9 - 93.5	5	4
A-	93.4 - 92.5
B+	92.4 - 91.5
B	91.4 - 85.5	4	3
B-	85.4 - 84.5
C+	84.4 - 83.5
C	83.4 - 77.5	3	2
C-	77.4 - 76.5
D+	76.4 - 75.5
D	75.4 - 70.5	2	1
D-	70.4 - 69.5
F	69.4-0	1	0

 

Ultimately both integer scales are converted to percentages for calculating a final grade across assignments, but those are spread out so that a failing grade doesn’t start until below 15%.  Guskey makes a point of pointing out that a traditional scale provides more room for differentiating different levels of failure rather than different levels of feedback for improving on those that are passing.  Let me clarify,  is it realistic that there are over 60 points to provide feedback of on failure and only 7 between each grade?  I’m not trying to say that each percentage must line up with a point of feedback, but using a percentage scale gives that illusion of precision.

• I understand that we may have rubrics that have lots of points for feedback.  It’s just easier for me to think of these points as they align to a specific grade or an outcome I am looking for.  In other words, “If you do these things that equates to a ‘B’.”  Rather than, “If you do these things, it equates to a 34 out of 45 or 86%.”

So where does this leave me?  It leaves me trying to circumvent that percentage system or complex points.  Why should there be a middleman between the grading of the assessment and the grade of the assessment.  

In the end, I think we all realize that points and letter grades only communicate so much.  It’s the qualitative/personalized feedback we give students that communicates the most.  Everything else is just a form of efficiency.  It’s important that we are never under the illusion that assigning more points or a greater span of percentages will ever give more precision than qualitative feedback.  

But it’s so hard to find that balance.  That’s why there’s tomorrow.