I’ve been asked this question by others, but unfortunately don’t have a good answer. I think the simplicity is a key feature, so extending it in ways that make it more complicated risks losing the effectiveness.
I had the same thought at first. I wonder if the additional dimensions that many of us include are really just subdivisions of “effort” (homework, attendance and engagement) and “competence assessments” (unit exams, project reports, lab reports, etc).
Thanks, Joshua! I love the clarity that your heatmaps provide. I'm going to experiment with this for my fall courses.
Here's a question. When I look at any of your examples, I am struck by how many more times F occurs than A does. The first example has 15 F boxes and 2 A boxes, if I counted right. This isn't a reflection on your grading approach, which is entirely reasonable; rather, it's just a visual artifact of the heatmap. It comes from the fact that when you consider all possible pairings of homework scores and unit scores, more of those pairings result in an F than result in A, which makes sense and is appropriate. But I'm wondering how students perceive this compared to the traditional table, which has equal representation of F and A (one row each). I could imagine a student looking at the heatmap and thinking "wow, so many ways to fail." Do you have any experiences with student perceptions on this point?
I haven’t heard directly from any student on this point, but I do think it’s worth thinking about. Actually, it used to be worse, because I was originally going to use the “expanded version” in my syllabus, until someone pointed out to me that students would probably have a negative emotional response to a charge that is 75% red F’s. So I shrunk it so that the red F’s take up less of the visual field. All this to say, I don’t have a great solution.
1. Very thoughtful visualization technique, thank you for sharing Joshua!
2. The key difference between weighted averaging and your standards based approach is thresholds: in SBG, there is a minimum requirement in both categories that students must meet to earn a particular grade. That is not the case for weighted averaging. The heatmaps are a very nice way of visualizing that difference.
3. I disagree with your assessment that weighted averaging includes a "bug" that incentivizes students to work on the easiest category of assignments. If those assignments are designed to support learning, why is it a bug? Also, if students can demonstrate proficiency on assessments, isn't it also a "bug" of the SBG approach that there is a threshold for homework to attain a certain grade? What if a student does well on assessments but doesn't do the homework? I'm not trying to be contrary, but asking questions to push your thinking. You're right that there is a fundamental difference in the grade mechanic between a threshold-based system like SBG, and a weighted-average system. I'm not convinced, however, that one is necessarily better than the other; a lot depends on the details of execution.
4. Thanks for taking the time to write this post -- it certainly stimulated my thinking!
Regarding you point 3, to me it's a matter margins, particularly towards the end of a course. As you say, assignments are designed to support learning, that is, we're all trying to assign our students work that benefit their learning. But from anecdotal observations of students in the final weeks of a course, they become more focused on grade outcomes around that time, and engage in behaviors which are geared towards that, like trying to get the last few points on an online assignment.
I've seen an unfortunate number of students get sucked into spending time optimizing for a few more homework points rather than investing the same time into finally learning some concept they never really mastered. I don't think SBG is perfect by any means, but I think it gently guides students a bit better at this phase of the course by making it more explicitly clear that they're better off spending time mastering some topic that'll be on the final rather than using a tool that allows them to get more online homework points without learning anything.
This was an interesting read. I am thinking of getting into standards based grading and I liked the heat map for communication. My question is regarding student understanding where they stand midway through the semester; is there a way for them to see that they are on track for A or B based on their progress so far in the assigned work?
The best I did for this in a recent course was to walk them through a calculation/estimate process halfway through. Basically I suggested that they take their current number in each category and double it, then look at the grade in the table corresponding to that. Obviously, this is a very naive and coarse estimate.
Cool idea. Any thoughts about how to extend this to three or four dimensions?
I’ve been asked this question by others, but unfortunately don’t have a good answer. I think the simplicity is a key feature, so extending it in ways that make it more complicated risks losing the effectiveness.
Seconding this comment. I like this a lot but my class has four categories of things instead of two.
I had the same thought at first. I wonder if the additional dimensions that many of us include are really just subdivisions of “effort” (homework, attendance and engagement) and “competence assessments” (unit exams, project reports, lab reports, etc).
Thanks, Joshua! I love the clarity that your heatmaps provide. I'm going to experiment with this for my fall courses.
Here's a question. When I look at any of your examples, I am struck by how many more times F occurs than A does. The first example has 15 F boxes and 2 A boxes, if I counted right. This isn't a reflection on your grading approach, which is entirely reasonable; rather, it's just a visual artifact of the heatmap. It comes from the fact that when you consider all possible pairings of homework scores and unit scores, more of those pairings result in an F than result in A, which makes sense and is appropriate. But I'm wondering how students perceive this compared to the traditional table, which has equal representation of F and A (one row each). I could imagine a student looking at the heatmap and thinking "wow, so many ways to fail." Do you have any experiences with student perceptions on this point?
I haven’t heard directly from any student on this point, but I do think it’s worth thinking about. Actually, it used to be worse, because I was originally going to use the “expanded version” in my syllabus, until someone pointed out to me that students would probably have a negative emotional response to a charge that is 75% red F’s. So I shrunk it so that the red F’s take up less of the visual field. All this to say, I don’t have a great solution.
1. Very thoughtful visualization technique, thank you for sharing Joshua!
2. The key difference between weighted averaging and your standards based approach is thresholds: in SBG, there is a minimum requirement in both categories that students must meet to earn a particular grade. That is not the case for weighted averaging. The heatmaps are a very nice way of visualizing that difference.
3. I disagree with your assessment that weighted averaging includes a "bug" that incentivizes students to work on the easiest category of assignments. If those assignments are designed to support learning, why is it a bug? Also, if students can demonstrate proficiency on assessments, isn't it also a "bug" of the SBG approach that there is a threshold for homework to attain a certain grade? What if a student does well on assessments but doesn't do the homework? I'm not trying to be contrary, but asking questions to push your thinking. You're right that there is a fundamental difference in the grade mechanic between a threshold-based system like SBG, and a weighted-average system. I'm not convinced, however, that one is necessarily better than the other; a lot depends on the details of execution.
4. Thanks for taking the time to write this post -- it certainly stimulated my thinking!
I'm glad to have stimulating thinking!
Regarding you point 3, to me it's a matter margins, particularly towards the end of a course. As you say, assignments are designed to support learning, that is, we're all trying to assign our students work that benefit their learning. But from anecdotal observations of students in the final weeks of a course, they become more focused on grade outcomes around that time, and engage in behaviors which are geared towards that, like trying to get the last few points on an online assignment.
I've seen an unfortunate number of students get sucked into spending time optimizing for a few more homework points rather than investing the same time into finally learning some concept they never really mastered. I don't think SBG is perfect by any means, but I think it gently guides students a bit better at this phase of the course by making it more explicitly clear that they're better off spending time mastering some topic that'll be on the final rather than using a tool that allows them to get more online homework points without learning anything.
This was an interesting read. I am thinking of getting into standards based grading and I liked the heat map for communication. My question is regarding student understanding where they stand midway through the semester; is there a way for them to see that they are on track for A or B based on their progress so far in the assigned work?
The best I did for this in a recent course was to walk them through a calculation/estimate process halfway through. Basically I suggested that they take their current number in each category and double it, then look at the grade in the table corresponding to that. Obviously, this is a very naive and coarse estimate.
Here's our post from quite a while ago about assigning midterm grades. It's more of a challenge, but doable! https://gradingforgrowth.com/p/midterm-grades