Active Grading: Scale Matters

A colleague of mine uses a 0% – 100% scale, but never gives anything below a 50% for any reason. So when a student answers that George Bush was the first president of the United States, he gets a 50%.  When a student doesn’t even hand in an assignment, she gets a 50% – not a zero.

Let’s picture it!

The traditional 0-100 scale:

The 50-100:

So clearly there’s an aesthetic advantage.  It’s just nice to have the grades spread out over our scale, isn’t it?  But there’s some mathematical beauty in here too.  In the first scale, a terrible grade averaged with a great grade makes… a terrible grade!  In the second scale, a terrible grade averaged with a great grade makes… a mediocre grade!  This is a huge selling point for me.  If you’re averaging your students’ scores, no longer do they have to earn THREE PERFECT SCORES to bring a single zero up to a SEVENTY-FIVE.

Also, I love the change in focus from the first scale to the second.  The first one has a huge red zone – so many scores that are failing.  The second one is mostly “you’re doing well” and “you’re getting there,” with only the room needed reserved at the bottom for “you’re failing.”

It feels a little weird giving anything but a zero for a totally negative assessment, but this is mathematically identical to making 50 the top score and 0 the bottom score.  Or having a 5-point scale:

If you’re looking for a way to progressify your grading scale a little bit and the rest of your school goes from 0-100%, or you’re mid-way through your class and you want to change a little without a lot of commotion… just stop giving zeroes!


This is cross-posted from the ActiveGrade blog.  ActiveGrade is the software I’m creating to support all of these new-fangled SBG ideas that make so much sense.  Though I’ll probably cross-post most articles that are relevant to the Point of Inflection themes, if you want to catch all of the updates to the program make sure to subscribe over there too.  Thanks for reading!

20 thoughts on “Active Grading: Scale Matters”

  1. I am intrigued by this idea and will have to give it some thought. However, my skin crawls at the idea that a student who just blows off an assignment can still earn a 50%…

    1. It definitely takes a redefinition. Right now you probably use 50% to mean “you got about half of the questions right” or something similar, and it would feel wrong to give a student that mark when they got zero of the questions right. The analogy between the 50-100 scale and 0-5 scale makes it easier for me – I think I could redefine 50 to mean “you got no questions right” and still feel like I was giving a zero.

  2. keninwa,
    I switched from the 0-100 scale to the 1-5 scale last semester. I give scores of 0, 1, 2, 3.5, 4, and 5. I still give zeroes for no work or for totally wrong work. Then I can multiply by 20 (and I tell students this, too) to see it in the traditional grading scale of 0-100. A 5 is 100%, a 4 is 80% (this was a little troubling at first, but with the ability to reassess, I was ok with “perfect” being my standard for an “A” equivalent), 3.5 is the next score down (I don’t give 3s) and that is a 70 % C equivalent. Anything below a 3.5 is not passing.

    This scheme was part of my way of easing into SBG in my community college calculus class.

  3. Before I Saw the Light and started SBG-ifying my gradebook I actually constructed an elaborate FileMaker database where I could set the “base” percentage for each “point-earning opportunity” separately. This was usually 50 or 60, as you suggest here. It was cool, it was baroque, it and it regurgitated plausible grades on a 0-100 scale at the end of the quarter, and it was impossible to explain to anyone; I’m so glad no one ever asked.

    Your excellent graphics make the whole problem clear, and help show why “90-100 is an A” is such a lousy paradigm. Thank you!

  4. I’m easing in to SBG this year with a 1-4 assessment scale that corresponds to a traditional grading system like so:

    1.0: Below Expectations (40% – F)
    1.5: Below/Approaching Expectations (50% – F)
    2.0: Approaching Expectations (60% – D)
    2.5: Approaching/Meeting Expectations (70% – C)
    3.0: Meeting Expectations (80% – B)
    3.5: Meeting/Exceeding Expectations (90% – A)
    4.0: Exceeding Expectations (100% – A)

    So 40% is the lowest possible assessment score, even if it’s just a blank page with a name on it (which has not happened yet).

    Here’s how I do homework:

    0: Assignment not handed in (0% – F)
    1: Few problems attempted (40% – F)
    2: Some problems attempted (60% – D)
    3: Most problems attempted (80% – B)
    4: All problems attempted (100% – A)

    So a student does get a 0% for missing assignments in my scheme…but I’m attracted to the idea of not giving zeroes as mentioned in the post. Homework is weighted at 20% of the student’s overall grade, so a zero here or there hurts (but doesn’t destroy) the overall course grade. Since I’m the only one at my school using an SBG system, this scheme seemed to be a reasonable compromise given the way my colleagues grade and the way the district’s gradebook software works.

    I plan on making more radical changes to my SBG assessment scheme for next year. In order for a student to earn an A (90%) on assessments for the semester, the assessment average must be 3.5 or higher and the lowest individual assessment score must be 3.0 or higher (indicating mastery of all course topics). ActiveGrade should make it easier to transition to such a system!

    1. If you don’t want to allow a student to miss an assignment, then it’s not about what your lowest score is. You should just immediately fail him when he misses an assignment. You could just refuse to accept any bad work. Then, you’re totally right: we don’t need to worry about giving a 0% or a 50% because none of that “crap” would even make it to our gradebook.

      After all, numerical grading schemes are largely a coverup for the fact that the system doesn’t leave enough time for real communication with and about students. But I don’t see anything about the number zero that makes it the one true score for the lowest permissible performance. If we ARE going to give students a grade that means “this is unacceptable,” I think 50 carries just as much merit as 0. You could use NEGATIVE 100 for unacceptable work if you really wanted to make it impossible to recover.

      I think of this level of design decision as more than tinkering. We should choose our grading policies intentionally!

    2. The problem here is the assumption in many %-based grading schemes that all assignments should be so easy that it takes getting 70% of the assignment right before you are barely passing.

      At the university level many faculty make exams and other assessments provide as much information as possible, so that the mean is 50% and the standard deviation around 20%. On these tests 50% is generally a B (most universities have enough grade inflation that the median score is B now, not C), and passing is around 30%.

      I think that you should resist the attempt to define A, B, C in terms of fixed high percentages, as that will set up unreasonable expectations in kids about how easy tests are supposed to be.

      Faking the percentages is simply serving as an enabler.

      1. Right. You could look at the idea of rescaling to the 50-100 scale the same way – it makes 75% the half-way mark, which you could hope would be the median grade.

        In any numeric scheme, I think it’s important to talk with students about what the numbers are supposed to represent. Everyone’s got a different system but all percentages look the same.

  5. My first school I in taught was an IB school that graded from 1 to 7, at least the last two years I was there. It was really rough transition for a lot of the teachers and students but we made it. The grade “1″ was really the only “failing” grade, 2′s were bad but in some cases the student could scrape by. In the IB grading system a 6 was pretty close to our A+. And the 7 was just really hard to get! In physics the students would get a 7 by roughly “earning” 73% of the points. It was great. It allowed so much room (or resolution) on the tests to really see what the students did and did not understand. Rather than writing 60-70% of the test as “gimme points.”

    I have been trying to push my current school away from the traditional grading. I haven’t got much of a response. But you idea is a mighty fine idea to adapt the traditional to something more progressive.

  6. I’ve done the 50-is-the-lowest-score thing as well with a few modifications…

    First, not turning something in is still a zero. (Actually, our grading software allows for “holes” that show up as “Missing” assignments when the progress report is looked up by students or printed by me). Occasionally I weirdly have students turn in blank work – this earns a zero. (yes, that’s weird, I agree)

    Second, When I hand back quizzes or tests I always give the kids their *actual* grade on the assignment. However, every three weeks when grades are handed back to kids (in our inner-city school we pass out grade cards every 3 weeks rather than the county’s 4.5-week schedule to give them more constant feedback) I’ll go through the gradebook and raise all the below 50% grades to a 50. *I don’t tell the kids I do this.

    We have a lot of discussions at our school about grading and I firmly believe that a kid who gets a progress report with a 53% (or whatever) sees that they can recover. However, a kid who earns a 34% sees no opportunity for fixing that grade – thus, they give up, slack off, and eventually get so far behind they’re a disruption to my class and my students.

    Let’s face it… a kid gets a 50% average, they’re still failing. But the difference is there’s HOPE. And sometimes, just knowing that they’re within range of pulling a grade up to passing is enough to get them to work and keep trying. Okay, I’m getting wordy. And I think I’ll take this to my own blog post! :)

    1. Thanks for sharing your experience. I’m in total agreement on the 53% vs 34% experience – they’re both failing, they both turn in to an “F” (or whatever), but one says “you’re failing and could stop failing” and one says “just stay at home.”

      Why do you record zeros if you’re just going to upgrade them to a 50% later? Why not tell the kids?

  7. I had a name for this: the “mercy rule.”

    I totally understand teachers who have a hard time justifying giving half credit for something that should be worth no credit, but that’s focusing on the wrong justification. Teachers need to focus on the capability of their grading system to reflect student ability and achievement. Once they do, it’s hard to justify the number of good grades it takes to negate the influence of a zero.

    1. That’s a fantastic way of putting it. This isn’t giving “half credit,” it’s redefining “no credit” and making it a position that’s possible to recover from. Thanks!

      I wouldn’t even call it the “mercy rule” – I’m tempted to call the traditional scale the “extra-mean rule!”

  8. I did this last year with my high school students at an alternative high school. Well, the lowest grade they could get on an assignment that was turned in (and not blank) was a 50.

    The problem was that the students turned in many mediocre assignments. When questioned they would respond that they got a 50 and that is close to passing. To them getting a 50 on mediocre work justified not putting much effort into the assignment.

    My question is: How do you explain this so the students are putting forth their best effort, not just trying to skate by?

    1. Sorry, I didn’t see this comment for a long time! It’s too bad your students turned to using this as a way to skate by. I don’t have any really useful advice, I’m afraid, but my gut feeling is that if they can get a 50% on assignments and still pass the class, they probably didn’t need to do the assignments anyway.

Leave a Reply