Don’t Give a Pass to Failed Grading Systems

Better late than never, this is my contribution to the work of a dynamic duo of fellow bloggers taking up problems with grading. Sandy Merz set the bar pretty high right out the gate with “Dear A+ Student” – and then he added the wonderful post Jesse Pinkman’s Wooden Box (love the Breaking Bad reference!). Brianna Crowley was next, arguing that grading is A Duct-Taped System in Need of an Overhaul. I couldn’t agree more. I hope you’ll check out their posts, comment, share this one, comment, and more importantly, take steps to help us change the norms and assumptions around academic performance and success.


 

Points are fine for basketball. We don't need them in grading.

Points are fine for basketball. We don’t need them in grading.

Considering the importance and consequences attached to student grades, in high schools especially, you’d think that we would have a broader professional understanding about what grades mean and how they work. Our consensus only goes as far as agreeing that grades usually represent some combination of knowledge, skills, and compliance. Beyond that, it’s a mess, and one that we ought to clean up.

Here are a few suggestions to improve grading practices. But even when we add them all up, I think we still have to consider whether or not we have a system worth saving.

Stop using the 100-point scale. Percents serve us well when we want to know relative quantities, and when the units themselves have a clear meaning. But using percentages for grading doesn’t pass that test. We think we understand the salient differences between students who post grades of 92% and 77%, but I’ve never met a teacher who could describe a meaningful difference between posting grades of 38% and 23%. What that indicates is that 15% doesn’t really have a consistent meaning in grades, in terms of the units making up the 15%, or the relationships among numbers in the lower half of the scale. In effect, we dedicate half of the 100-point grading scale to punishing children, creating mathematical levels of failure with no meaning in reality. An objective analyst would conclude that such a system is designed to make improving from a 20% to a 45% just as hard, and as significant, as moving from a C to an A+.

But – if your top grade must be 100, get rid of zeroes. Some teachers freak out when district mandates tell them they must use 50% as a floor and not assign grades below fifty. I totally get it. Teachers don’t want grades that seem to indicate “half” when the truth is less than that, or even “none”. (Briana Crowley’s blog post describes a conversation where teachers get caught up in that very issue). However, if we are committed to having zero mean zero, we still have no real assessment-related use for 10, 20, 30, or 40 out of 100. A grade scale of 0-50 would be more logical from a measurement standpoint, but then teachers would have to get over the idea that a grade of 25 on a 50-point scale would likely represent a grade of C. Of course, it all depends on what those numbers are counting and measuring. Suppose we ask students easy, medium, and difficult questions on a topic. If they answer all of the easy ones correctly and manage a portion of the more challenging questions, perhaps a grade of C would be appropriate. On the other hand, if we’re counting uses of welding equipment without injury, then even 45 out of 50 is unacceptable. We need to relinquish our preconceptions about the meanings of specific numbers and percents. Giving up the idea of points altogether would help; points are a convenient fiction, as long as you don’t think too hard about what they supposedly represent.

Convert grades to a different scale. Another way to look at it is this: we have five letter grades in a typical system (A, B, C, D, F), and in some cases, we use plus and minus signs to differentiate within those grades. Assuming you don’t have F+ or F- you’d end up with thirteen possible grades. Why do we need 100 points then? That’s a level of definition that has no meaning. It would be like having a weather report stating today’s high temperature was 58.3 degrees, or including cents in conversations about rents or mortgage payments. So, use points and percentages if you must for an individual assessment, but convert the assessment score to an appropriate scaled grade (like A-F, or a thirteen-point scale matching letter grades with plus/minus marks, or a 0-4 scale possibly using 0.5 increments).

Stop using averages that distort learning. If you studied music or martial arts in a system that averaged your assessment scores, you’d never be able to achieve musical commendations or a black belt in your sport! Averaging has the effect of penalizing students for early performance, before learning occurred. If a student couldn’t solve algebraic equations well in September but has mastered them in December, why are the September failures still worthy of grade consideration?

Try a standards-based approach.  Standards-based grading encourages teachers to assess carefully, and encourages students to continue striving for mastery without the dead weight of prior struggles holding their grades down. Teachers have to know exactly what they’re assessing for this to work; instead of grading categories like tests, homework, and classwork, we need categories that fit standards in our curriculum. If you must assign a letter grade in the end, then take some kind of average of demonstrated skills across standards; at least then you’ll know that the grade is more of a reflection of eventual mastery rather than a muddled picture combining past and present.* To the extent that compliance does matter in student grades – turning work in on time, not cheating, etc. – but those grades in an appropriately identified category. I call mine, “Student Skills” – and since typical performance matters in student skills, I do average those grades throughout a grading term.

In practice, this means that complex assessments like essays and projects generate grades in multiple categories. It sounds confusing, but actually generates only an extra few marks and keystrokes for the teacher, and more importantly, it generates more thinking for students. Whereas my students used to have a single grade of A, or B, or C on an essay, they now receive separate marks for writing skills, reading skills, and student skills. Keeping the grade elements distinct means that students (and parents) can see at a glance which skills in particular are strongest or weakest. This approach also helps mitigate some of the complacency that occasionally accompanies an A or a B, or discouragement that comes with a C or lower.

Ready for a really bold change? How about throwing out grades? There are teachers out there doing it, and they offer some compelling reasons. Find them on Twitter (#TTOG for “teachers throwing out grades”); look at blog posts by Starr Sackstein, and the work of Mark Barnes. If you’re not ready to go that far, look up assessment and grading practices information Robert Marzano and Doug Reeves, and check out #SBG on Twitter. Whatever you do as a teacher, don’t shy away from the discussion.

* Sentence revised for clarity on 3/2/15.

12 thoughts on “Don’t Give a Pass to Failed Grading Systems

  1. I’d like to give another grading system a try. I wonder if a standards-based system off of the CCSS Mathematical Practices would work in place of standards? I like your idea of a “student skills” category. Is it weighted the same as the other standards?

    1. Thanks for reading and commenting, Susan. Regarding the student skills component of the grade, I think it would be appropriate to weight it differently depending on the grade level and type of class. For 9th graders, I think a little more emphasis on those skills is appropriate, while for seniors, I might reduce it a fair amount. Perhaps the same thing for regular vs. honors classes. But once you get down to those decisions, you’re pretty far along already, I think. The most important thing is to maintain clarity about what you’re assessing, why you’re assessing it in a certain way, and how that assessment fits our goals for student learning.

  2. I’ve been tinkering with formative and summative assessments this year, and the jury is out. Because most test scores are formative and therefore “don’t count” in students’ minds, students seem more focused on gaining skills and knowledge. Especially for my AP kids, this system disincentivizes cheating and relaxes this sometimes anxious group. However, now that I’ve had four grading periods to review, my grades are pretty much the same as they’ve always been. It seems counterintuitive, but there it is.

    1. Thanks for sharing your experience, Vicki. I think the difference in attitudes and cheating are great benefits of this approach. As for grade distributions, I have a slightly different experience, but haven’t taught any AP classes, or even any honors classes above tenth grade. I’ve found that most students who would have earned D’s and F’s in the traditional system can manage a C in this system (most, not all). The reason won’t surprise any experienced teacher: we know how often D’s and F’s are a reflection of work habits or life circumstances rather than skills. And some students who would have had C’s in the old system get into the B range. But I have not seen a huge spike in the number of A’s – and that makes sense. Students who earn A’s in either system tend to be those who master the material and demonstrate good habits and “student skills.” What we’re looking for from our students hasn’t changed that much, but a different grading system makes feedback more transparent, and offers a better fit in terms of what we know about motivation and the learning process.

  3. Hi David, thanks for this post. I work in a Thai school which is very much focused on summative exams & grades-based. However, there is also a large coursework & project component, and I found some helpful ideas here which I will be able to incorporate into my approach to grading.

    1. Great to know that this was useful, and even more so across the world! Thanks for reading and good luck!

  4. “In effect, we dedicate half of the 100-point grading scale to punishing children, creating mathematical levels of failure with no meaning in reality.”

    Excellent point!

    Any system that is built for system wide comparison of school performance will never be useful for classroom based assessment, that serves the purpose of understanding individual students needs and modifying practice to serve them.

Leave a Reply

Your email address will not be published. Required fields are marked *