Below you will find an outline of my current grading policy and the reasons behind it. If you have any questions or comments about what you see here, please send me an email at Thomas.Gribble@tusd1.org or fill out the contact form on this website. I will also be maintaining a blog throughout the 2021-2022 (maybe 2022-2023 also!) school year about how this grading policy seems to be working, and you can view that blog here. I appreciate your patience and open-mindedness as I experiment with this system and try to come up with the most equitable grading policy I can!
My grading policy this year is based on two main elements: category weighting and a new grading scale. The two categories in the gradebook are Assignments and Summative Assessments, where Assignments are anything that is formative, and Summative Assessments are anything that is summative. Assignments are weighted at 0% of the final grade, while Summative Assessments are weighted at 100% of the final grade. We will also not be using the traditional 0-100 grading scale (percent scale), and will instead be working with a 0-4 scale, similar to that of the GPA scale. A more thorough explanation of these two elements can be found below.
In the Spring 2019 semester, I changed the grading policy for my Honors Precalculus classes to a "homework can only help policy" in response to my findings from analyzing the data from the previous semester's Honors Trigonometry grades. I made two categories in the gradebook: one called "unit score", worth 80%, and one called "projects", worth 20%. The unit score consisted of points from things like homework, in-class assignments, review activities, and tests. I sold doing homework as something that benefits you in two ways: as a mathematical buffer against poor test grades, and as a tool to help learn the material. If students had a better score on the test than they had on the homework, their homework score would not count for that unit. On the other hand, if students did very well on the homework and did poorly on the test, then the homework score would serve to boost their overall grade on the unit. If students did not do the homework at all, then their grade in the unit score category would be based entirely on their test scores. There were three mandatory projects that semester that students mostly did well on.
At the end of the semester I was interested in diving into the data to see what impact, if any, this policy had on student grades. This was far from a controlled experiment, but my takeaways were basically this:
There was not a significant correlation between test scores of students who did the homework consistently and those who did no homework.
The "buffering" effect of the homework scores was much less significant than I anticipated.
Taken together with feedback from my students, these findings were enough for me to implement the same grading scale across all of my classes the following school year. Throughout the semester, students reported feeling less stressed about their math homework because they knew it wouldn't count against them if they were unable to turn it in due to increased time commitments in other classes or extra curricular activities.
During the 2020-2021 school year, and to some extent the 4th quarter of the 2019-2020 school year, I observed a few things that made me want to investigate my practices. While remote learning was difficult for almost everyone, it was more difficult for certain students: those who did not have the same time and resources as others. Although it is 2021, we have to remember that not everyone has the luxury of computers and internet connections at home. While the district attempted to make sure that technology limitations would not be the reason students could not succeed with online learning, there was still a lot of inequity among students.Â
I don't know if I was explicitly told this, but I was under the impression that during remote learning, my grading policies had to be very similar to those of teachers with whom I had shared preps. The inequities described above produced some of the lowest grades (percentage-wise) I had ever seen in a gradebook. I had students who had single-digit percentages in my class: not only had they done very little if any homework assignments, they were also doing poorly on tests. It is easy to rationalize that students are doing poorly on tests because they hadn't done any homework. My findings from previous years indicated that this was not always the case. After the semesters had neared their end, I did my best to implement some adjustments to grades to help out students who had done poorly, but several cases were too far gone. This left me feeling like I had somehow failed my students - like I had failed to provide them with the opportunity to be successful.
After that difficult school year I promised myself that I would look into making significant changes to my grading practices to prevent this kind of thing from happening. I had some suspicions about why this might have happened, but while researching over the summer, the reasons became quite clear.
Everyone can agree that grading should be equitable.
The problem is that very few teachers were ever taught how to grade, and grades are one of the last bastions of autonomy that teachers have. Personally, although one of the classes I took during my Master's program in Teaching and Teacher Education talked a bit about how grades should never be used as a behavioral consequence, there was never a discussion about how to properly set up a gradebook and evaluate students. There was never even a clear definition of what an "A" was, or a "B", or a "C", etc. I believe the masters program I went through was fantastic and I have nothing negative to say about it, but grading is simply not something that teachers usually receive a formal education about.
Rather, I think the way we grade (at least in the beginning) is a product of two things:
Whatever our "mentor teacher" (the teacher we student taught under) did.
What our own teachers did when we were students.
Furthermore, because grades are one of the last things that we have "the final say" on, teachers can feel very defensive of the way they grade (that was inherited), even when presented with evidence of its flaws (that were inherited).
So what is equitable grading?
It is worth noting that most of what I am doing this year is based off of ideas I read in a book called Grading for Equity by Joe Feldman. I do not read a lot so this might not be compelling, but I cannot possibly recommend this book enough for teachers, administrators, parents, and students who are interested in this topic. According to Feldman, equitable grading practices must have the following characteristics:
They are mathematically accurate, validly reflecting a student's academic performance.
They are bias-resistant, preventing biased subjectivity from infecting our grades.
They motivate students to strive for academic success, persevere, accept struggles and setbacks, and gain critical lifelong skills. (Feldman, 71)
I will attempt to summarize what each of these characteristics entail.
If grades are supposed to describe a student's level of academic performance, we must accurately calculate them in a way that is easy to understand. While most people understand what an average is, they might not realize how inaccurate a representation of student performance calculating an average can produce. There are two reasons for this. First, of the three measures of central tendency, the mean (average) is most heavily impacted by outliers. If a student gets a zero in the gradebook, it can become nearly impossible for them to overcome that. This is not a matter of a student needing to work extra hard to climb back from a zero, rather a mathematical obstacle that needn't hamstring their potential for success in the class. Second, when we simply average scores across a term (most gradebook software's default and only behavior), we are unable to account for a student's growth over that term. If a student earns full points on something towards the end of the term, indicating mastery, while they had previously done poorly on a prior assessment on that topic, an average does not account for the more recent evidence of mastery.
There are many things that can be done to fortify the mathematical accuracy of grades. One set of measures we can take involves mitigating the effect of outliers on the average calculation. The reason a zero is so destructive to a student grade is not because it is a zero, but because it falls at the very bottom of the traditional 0-100 percentage scale, which has "desirable" scores heavily skewed towards 100. When 59% of your grading scale represents failing, giving a 0% on something becomes almost cruel. Consider a student who is given three equally weighted assignments and gets an A on two of them but for whatever reason does not turn in the third. This student has demonstrated some form of mastery on the two assignments, but if we consider their A's as 95% and the missing assignment is scored as a 0%, then their average on those three assignments works out to be a 63%, D. If a student continued to get these assignments, they would have to get an equivalent A (95%) on 15 straight assignments just to bring their average to 90%.
We accomplish this mitigation of the zero by redefining a grading scale to be more distributed across the point spread. For example, we can adopt a 0-4 scale, similar to that of a GPA scale, where 0=F, 1=D, 2=C, 3=B, 4=A. That same student who earned an A on those two assignments would have a very high C, and would need 5 more A scores in order to have an A average (assuming >3.5 is an A). This is a good start. We can accomplish roughly the same thing without discarding the percentage and traditional 0-100 scale by adopting minimum grading, which redefines a zero as a 50%. It may seem strange to not give a zero on an assignment that was not turned in, but a 50% is still an F. We don't need to be telling kids that they have "super failed" by giving them a zero.
Other measures we can take to insure mathematical accuracy of grades include things like weighting more recent performances more heavily, or even allowing more recent performances on similar topics to completely replace past performances*. (*If gradebook software does not have this functionality built-in, it can be tremendously difficult for a teacher to keep track of this kind of thing.)
Grades should be based on valid evidence of a student's content knowledge, not evidence that is likely to be corrupted by the teacher's implicit bias or reflect a student's environment. For example, if a teacher were to grade on participation in a class where 100% of the content is not participation based (like P.E. or art, for example), then they would be creating inequity by making all students adhere to some pre-conceived idea of what good participation consists of. Some students learn best by observing others, while some students are very shy and have a hard time working in group settings. If those students are still able to demonstrate strong content knowledge, then there should be no penalty for their lack of "standard" participation.
Cultural differences also exist between students, which makes giving grades on anything behavior related inequitable. Some cultures engage in spirited discussions at home that could seem out of place in a classroom, but if points were taken from a student for an "outburst" that is really just a culturally appropriate response for them, then that would be a problem.
We also need to consider that not all students' home situations and other obligations are equal. If a teacher assigns something as extra credit to be done outside of class, then there will inherently be certain students that will be more able to complete this task than others. Giving a grade based on an extra credit opportunity is inequitable. This same idea can apply to homework.
When students have more time available to do school work, they are going to be more able to do homework. As we know, not all students have the same amount of time available in their day to work on things outside of school. This means that homework assignments, or rather grading them, are inequitable. Not only this, but homework is something that teachers would agree is supposed to be part of the learning process. Giving a grade on something that occurs during the learning process (formative), is potentially punishing a student who has not quite been able to process and learn all of the information yet. For these reasons, grading homework (even for completeness, but especially for correctness) is inequitable.
Grading that is resistant to bias should:
Be based on required content, not extra credit
Be based on work, not timing of work (no grade penalty for late work)
Be based entirely on summative assessment and not formative assessment
Not be based on participation or behavior
Additionally, teachers should consider alternative consequences for instances of cheating. My personal belief is that if the incentive for cheating is removed, then cheating will naturally stop. However, if a student cheats on an assessment, giving them a zero is not an ideal punishment, because it sends the message that you are OK with the student not learning that material, and the student has no motivation to go back and learn the material because they can no longer receive credit for it.
We should strive to make grades motivate students to achieve academic success and supportive of a growth mindset. When we do things like not allow late work (or penalize late work), hold past poor-performances against students despite more recent demonstration of mastery, and similar things, grades become punitive. If we move away from a punitive grading system or one that awards points for tasks, then students will be more likely to believe that they can achieve success in the class. In order to make grades motivational, teachers should:
Give opportunities for redemption (retakes)
Make their grading policies and procedure transparent (students can hit a target that isn't moving and that they can see)
Emphasize constructive feedback
Adjust their language to emphasize learning over scores and grades
Promote soft-skills such as self-reflection, metacognition, etc.
Assignments that students complete at home or in class - things such as homework, group activities, and mini quizzes - go into the Assignments category of the gradebook that is weighted at 0%. As outlined above, there are a couple of reasons for this. First, these are things that I am asking students to do WHILE I intend for them to still be learning the material. I would not expect to be evaluated on my ability to solve a Rubik's cube while I am still learning how to solve a Rubik's cube, so I don't expect students to perfectly understand something while they are still learning. Giving a poor score during the learning process also sends the message that it is not OK for students to make mistakes. Second, I recognize that my students have wildly different schedules, where some of them may simply not have enough time in the day to complete whatever homework I do assign. If that student is able to improve their understanding of the material by doing less of the work, or if they are able to understand the material simply through what we cover in class, then I should be OK with that.
Although these assignments are worth 0% of their grade, I still expect students to do them. I do believe that things like homework and practice play an important role in the learning process, I just don't believe all students need to do the same amount of practice in order to master a concept. I record scores for these assignments in the gradebook, and a student can turn any of them in late with no deduction. These entries in the gradebook serve as data points for the student, the student's family, and myself about how the student is doing in the class. If a student has many missing assignments and has done poorly on assessments, then we can start to address this by making sure the student is doing their assignments.
While things like homework, in-class activities, and quizzes are all formative assignments or assessments (things that are meant to be done during the learning process and reveal where a student is at with their current understanding), summative assessments are anything that I have a student do AFTER the learning process is meant to have finished, and go into the Summative Assessments category of the gradebook that is weighted at 100%. The most commonly used summative assessment would be a traditional math test that students are familiar with. There are also other types of assessment that can be summative, like projects, performance tasks, verbal "math talks" and more. Retakes will be offered for all traditional summative assessments (unit exams). If I want a student's grade to truly represent their content knowledge, then it should not matter if a student didn't quite know something at first but then learned from their mistakes and is able to prove their knowledge at a later time. That being said, retakes are the most logistically difficult part of this grading policy for me. Here are some details about retakes.
Traditional summative assessments (tests) are divided up into "content strands", so that problems that are meant to assess related topics are grouped together. This loosely resembles dividing a test up into standards. Students can approach each strand in sequence or hop around the test, but for each strand I ask them to rate their confidence level based on our in-class 5-1 scale [5 - God tier, 4 - Got it, 3 - Getting it, 2 - Wait..., 1 - What?]. Students receive a score from 0-4 on each section. This score is rubric-based. I am still tweaking this rubric and will publish it here later, but basically: a 4 means it is clear the student completely understands the topic that was being assessed (note, this means a student can get the wrong answer and still get a 4 (not heavily penalizing silly arithmetic mistakes)); a 3 means that the student has a fair amount of understanding of the material and made a topic-related mistake that lead to an incorrect answer; a 2 means that the student did demonstrate some understanding but not enough to correctly provide a reasonable solution; a 1 means that the student has put some relevant ideas down on the page; and a 0 means the student did not make a reasonable attempt on the problem or what was written did not indicate an understanding of the material. The scores from each section go into their own column in the gradebook.
If a student would like to retake the assessment, they will only need to retake the sections they did poorly on. There is no justifiable reason that I can think of to have the student retake an entire assessment if they just did poorly on a single section. In order to earn the retake, a student must complete a process that involves completing this form. This form is designed to have a student reflect on their performance on the original test, go back and correct their mistakes, and think about what they would do different to prepare next time. Upon completing the form, the student will bring it to me and I will review it and talk to them about what is on it before we schedule a time for them to retake the sections of the test they want to retake. This usually happens before school, during lunch, or after school. On some occasions this may happen during class time if it is the most effective use of our time. A student's score on the retaken sections will completely replace the original section scores and no penalty or "limit" will be assessed. There are also no restrictions to who can do a retake. A student who only barely missed one problem is just as entitled to a retake as a student who missed every question.
I am still struggling with deciding on a limit to how long after the original test a student can do a retake. I think idealistically I would not have a limit, but for my sanity and organization there simply needs to be one. If you can think of a way for me to do this more effectively, please reach out. (Update 6/5/2022: There definitely needs to be a limit because being flooded with retakes in the last one or two weeks of school is way too stressful.) Overall I am worried that this retake policy is going to create mountains of additional work for me. But that's OK I guess because I am a teacher and only work 7.5 hours a day for 180 days in the year??? (/sarcasm)
In the mathematical accuracy section above, I laid out how the 0-100 scale is garbage. This year I will be using a 0-4 scale. I will still be giving zeros for missing assignments or lack of work, but because those zeros are no longer extreme outliers and because the assignments category of the gradebook where zeros would (mostly) show up is worth 0% of the grade, they will no longer prevent a student from being successful due to the gradebook software's default behavior of taking an average.
The way the 0-4 scale is broken up for the 2023-2024 school year is as follows:
A: 4-3.5
B: 3.49-2.75
C: 2.74-2
D: 1.99-1
F: 0.99-0
This is the same scale I used for Semester 2 of the 2021-2022 school year. I wrote about the rationale behind this in a blog post. For posterity I will keep my previous thoughts about the grading scale and updates posted below.
**NOTE: EVERYTHING FROM HERE DOWN TO THE START OF THE CONCLUSION SECTION IS OUT OF DATE**
There are a few things I am still working out about this grading scale. The primary issue I am having right now is deciding where to have cutoffs between the different letter grades. I would like very much to not have to give letter grades entirely, but since that is not realistic, I would at least like those letter grades to be representative of student knowledge. I am not fully convinced that my current cutoffs (4-3.67: A, 3.66-2.67: B. 2.66-1.67: C, 1.66-0.67: D) accomplish that. My biggest issue right now is that based on how I have been grading assessments, a score of under a 2 should NOT be a C because it sends a message to the student that everything is OK with their performance in class, when it most certainly is not. If I find a solution to this and remember to come back and update this page, I will. (Update: I made an adjustment both during the first semester and again at the start of the second semester. See blog for details.) I may also start to enter individual "content strand" scores into the gradebook after a summative assessment rather than an averaged score. I think this would accomplish more clarity in what a student has or has not demonstrated mastery of, and will potentially allow me to more easily override scores on old performances on similar content with more recent ones. (Update: I did this and it was way better.)
**UPDATE 10/1/2021**
I have changed the cutoffs mentioned above to (4-3.50: A, 3.49-2.50: B, 2.49-1.75: C, 1.74-0.75: D, 0.74-0: F). I am not convinced this my final revision to this, but here's why I did it. First, I thought the band for an A was too narrow. Many students were doing great in my class and were sitting at 3.6 or so and that was calculating to a B in the previous cutoff setting. Second, I thought the band for a C was too wide. A handful of students were doing quite poorly in my class and were sitting at a C. Because of how the traditional grading scale works, those students may have been feeling complacent with that. On my scale a C is not meant to be a grade students should be comfortable with, so I took ~.16 out of C and put it into D.
So that's about it. I think what I have produced for this year is a fair and equitable grading policy. If you disagree I would really appreciate hearing from you because I know that other viewpoints exist and at the end of the day, none of us truly know what we're doing. Please don't hesitate to contact me through my email address Thomas.Gribble@tusd1.org or through the contact page of this website. Thank you for reading.