Grading for Equity
gse.harvard.edu
gse.harvard.edu
The problem with that, is that for the student who does well from the very beginning and gets A's on everything, their performance is fine, their average is an A, but for the student who struggles at the beginning and gets very low grades, D's and C's and even F's as they are in the process of learning, and even on early quizzes when they demonstrate mastery on the test and let's say they get an A on the test, because they have those earlier grades that ostensibly were for assignments and assessments that were on the path to learning, that they were supposed to learn from, and that they weren't even supposed to have learned everything yet, when we include those early scores, it pulls down the final grades, so it actually misrepresents the level of mastery that a student has ultimately demonstrated.
I tend to be skeptical of the term "equity" since I strongly disagree with a lot of the thinking that goes along with it, but I can empathize with the above statement. Anyone who works with data knows that using an average can product highly misleading numbers, so he's correct on that point. But also the idea that a student is supposed to be learning and improving, and that expecting A's across the board defeats that purpose, seems totally reasonable.
No system is perfect and there will be inadvertent and undesirable incentives to underperform one way or another, but I like the idea of a system that focuses on rewarding improvement.
For example if current grading practices are based on merit, then it is not equitable, as some students will naturally have higher grades than others. The solution, then, is to change the grading system so that good grades may be "redistributed" to lower performers and achieve a more "equal" outcome.
That's interesting, because this idea is so completely counter to what is proposed in the linked article. The article, if anything, is advocating for grades to be more focused on actual learning outcomes (performance) than is the current default. Many of the solutions proposed (such as getting rid of averaging over the entire course) have the goal of making it based more on merit, not less.
I think this gets to the difficulty, in terms of the overall discussion about race, equity, etc. There are some people who really abuse these terms, and others who simply don't know what they mean, or simply have very different ideas of what terms like this mean. (I see this from people with a variety of ideologies and backgrounds.)
My general conception of equity is about adjusting the system in order to remove unfair (and un-necessary) burdens on certain groups of people.
For example, equitable access to real food requires actual groceries available to people. I used to live in Park Slope, Brooklyn. There were literally five high quality grocery stores within a ten minute walk. Contrast that to other parts of the country where there is no easy access (at all) for people to buy fresh produce. That's not equitable.
Similar for grading: If you go into an algebra 101 class, and you do very badly in your first few weeks, simply because you have a lot to learn, you should still be able to get an 'A' in the class if you fully master the material. Otherwise, it's not equitable.
Equity specifically isn't (and shouldn't) refer to equality, IMO. It should be about removing the often arbitrary burdens that disadvantage some people, and the unfair advantages that only serve to put people in an advantaged position (such as providing better grades to people who already know the material in a class, when final learning outcomes are no better.)
So to me it seems we have a definitional issue with the word. Your interpretation of equity makes sense, and is not something I would be opposed to personally. But then there's the other interpretation (redistribution), which is shared by enough people that it's what I think of when I hear the term.
I don't think there's a right or wrong answer to "what equity means", and it's not a very interesting question anyways. What is interesting is how these two very different ideas became conflated, and how to move forward with the good idea and discard the bad idea.
I suspect they've become conflated in some circles because redistribution is seen as a real solution to lack of equity. Notably, this is literally baked into our system in terms of federal income tax, but many people argue that it doesn't go far enough. For example, redistribution of wealth could result in making childcare available to most people, eliminate childhood poverty, etc.
I don't think that in principal that 'redistribution' is controversial. Every contemporary society has some form of redistribution of wealth. However, many people see proposals around it as extreme. Though, whether the proposals are extreme may depend on one's knowledge of the consequences of action (and inaction), among other things.
This gets sticky, because a lot of social activists get hung up on this to the detriment of making real-progress on things that are not controversial when presented in the right way. I see a lot of calls for "ideological purity" and "absolute dedication to the cause". These calls, among activists, often become central to the conversation, as opposed to specific policy proposals that could make a real difference. Personally, I find it rather upsetting to see this happen, as I really do see the problems that are caused by lack of equity, structural racism, etc. But the way the dominant activist groups push for change seems to be too focused on ideology, as opposed to pragramatic solutions. I see the value in anger, which turns into dedication and motivation, I think that it sometimes goes way too far, and then becomes a form of unfortunate self-sabatoge. Anyways, I think I've digressed a bit.
edit: I should add why I tend to be skeptical of them. It's not just about idealogical purity, it's that in order to truly make progress, you have to reduce the total number of problems in the world. It's very easy to spot a problem - anyone can do that. It's less easy, but still fairly easy, to recommend a fix for the problem. But what's very challenging is to find a fix that does not cause more problems than it solves. It's entirely possible to make things worse, and the best way to make sure that doesn't happen is to be dispassionate. But right now the trend is to lean into passion, and specifically, anger. As you said, anger can prompt action. But the problem with anger is that it's a fire that you can't control, especially when you're talking about millions of people.
Critical social justice is a portmanteau term coined by Pluckrose and Lindsay that is being heavily pushed by media sympathetic to their brand of culture war. Neither of them are idiots, but equally I don't find either of them to be highly intellectually honest. I recommend avoiding the term CSJ on HN except where we want to talk about Pluckrose, Lindsay and their entourage, since I think the term tends to undermine curious conversation.
The article is strongly about more performance-oriented, meritocratic grading, not less so.
It considers this to be a problem of equity since current grading practices (in the opinion/experience of the author, at least) disadvantage certain students who perform better than others but have other problems (less access to tutors, less time for homework, less preparation before class etc).
To explain it succinctly, I've become wary of a certain group of people attempting to use subtle language choices to slip in their controversial ideas without being noticed. Equity sounds very much like equality, and therefore people are more likely to sign up for it without understanding the difference. It's a weasel word.
On one end of the spectrum you grade work throughout the semester, which has the failings you mention above.
On the other end a “final” of some sort is worth 100% of the grade, which feels too high stakes to be a reliable representation of what students truly understand.
I hope to find an explicit recommendation for a good balance.
https://styluspub.presswarehouse.com/browse/book/97816203624...
Alternatively, one could keep their existing grading system and simply use some sort of quadratic weighting scheme which continuously increases the effect of an exam or assignment on a student's final grade as the end of the semester approaches. This would be a lot easier to write down on a syllabus but might disincentivize effort spent earlier in the class.
Despite the title, the interview is arguing for grading practices that represent mastery in the class as precisely as possible. They are arguing for meritocracy.
Some specific proposals include:
- not grading homework as it is universally copied - not using grading to maintain class discipline - not giving extra credit for conscientiousness as it gives students a skewed view of their actual mastery of the subject matter - not using grade averages as the final grade for a class, as it doesn't reflect correctly the final knowledge that a student has accumulated by the end of a class (if they are a slow learner)
The author does believe that all of these practices are particularly disadvantageous to certain students with "bad backgrounds", hence the equity angle. But these practices are in fact not helpful for any student who actually wants to learn - they are just as disadvantageous to privileged students as well, as they inflate their grades above their mastery, and set them up for disappointment later on.
It is true that this is not a proposal to "get rid of grading" or "pass everyone," but the simple fact that this is trying to maintain meritocracy and improve learning outcomes is the bare minimum and does not mean that it is beyond reproach.
I recognize that this is an attempt to merge equity and meritocracy, but I disagree with the approach and rigor of the method. I don't think it's the obvious improvement that you might hold it to be.
I’ve been inspecting the book this week and agree with many of the proposals (don’t grade homework, for example). The one I take most exception to after a first pass is the changing of the grading scale.
But still reading it. Shared the link here to get some other perspectives.
I was just commenting based on what I had already seen in the early comments: many people saw the title and assumed that the author is proposing something like boosting the grades of minority students, and started arguing from there.
Unfortunately, "equity" is a really loaded term in today's American discourse, and I believe the author made a huge mistake in using this word for their proposal.
It is academic fraud to mislabel student grades.
It cheats future employers who will stop accepting degrees as a measure of quality.
It cheats students who rightfully paid and honestly earned their grades.
Most of all it treats the students who didn't earn their grades as less than everyone else. If they were not adequately nutured into a quality student, they should be tutored and assisted in getting up to standard. Consider Always Sunny in Philadelphia, does it help Charlie when he is told he is a genius?
We have this problem right now with anyone who has a sports scholarship. It can be assumed they didn't earn their degrees the same way a non-sport student did.
Fortunately, the author is proposing exactly this. Their critique of current grading practices is precisely that they do not correctly reflect student performance.
The "default" for most grading software is to average the score of all homework throughout the entire class. The result of this averaging is that people who actually learn the most might end up with a worse grade than people who learn the least. Why? Because people who struggle in the beginning, get their poor results put into their grade, even if they have complete mastery of the subject by the end of the class. This is clearly not ideal. A grade should be based on the level of mastery of the subject at the end of the course, and not at the beginning, when you haven't yet had the chance to learn the subject.
This, of course, could be changed by simply changing the default behavior of grading software. Most people use the defaults, so the software developer really does have quite a lot of influence.
It is prioritizing all the wrong things. If grading is inefficient then it ought to be improved, but replacing qualified achievement with instant gratification is very sad to see and will only harm the education of all students subject to the practice.
I don’t think we actually see the outcome you are referring to, where students who struggle at first end up with the same level of mastery. I don’t think we should use the grading process itself to compensate for inequities if they do exist.
please tell me something that is perfect in the world, just one of it.
education equity(read: score equally) is popular from k-12 now onto college, seems nothing can stop that.
path to hell is paved with equity "good-will" thoughts.
The article is arguing precisely the opposite: focusing grades more precisely on student performance, instead of extraneous things like time dedicated to school, discipline, conscientiousness etc, that are all a significant part of grading today.
For example, the author (interviewee) points out the practice of adding extra credit for conscientiousness, which makes it exactly that students are graded incorrectly compared to their actual knowledge on the subject.
In another example, he points out how students who start out doing poorly in a new class but by the end have improved and have absorbed the knowledge will nevertheless get a mediocre final grade because their final grade is an average.
In a third example, he points out how homework is often copied or done by others, and so again, grades for homework tend to misrepresent a student's actual knowledge in the class.
No, the author wants final grading to reflect ultimate subject mastery, not (e.g.) be distorted by details of the process by which that mastery was approached. Part of this is reducing demographic disparities, but only those demographic disparities in grading that are artifacts of the system grading for things other than ultimate mastery (including, as the article notes, giving what amounts to bonus on the final grade for coming in more advanced even when there is no difference in final mastery.)
> Producing "equity" as in equal results is not their job.
“Equity”, as used here, is not equal results independent of mastery. It is equal results for equal mastery.
Can you point out where this is expressed in the transcript? I am missing it.
What the guy specifically said has nothing to do with the political "equity" or "inequity". The points he raised are issues common in educational measurement to get more valid scores. It is pathetic that he tries to spread his voice by appealing to a political side.