Welcome to Science with Shrike! Today we’re going to discuss one of the more controversial parts of the holy DEI triad—equity. While this is a relatively new add to the acronym, the concept of equity has existed as a core challenge of teaching, and measuring teaching success, long before it was applied to minorities or adopted into the DEI trinity. This pre-existence may account for the widespread popularity of this idea, why many teachers are quick to adopt it, and why banning DEI will not kill it.
What is Equity?
Out of “Diversity, Equity, and Inclusion”, “Equity” causes the most confusion. Diversity—we want more minorities in our program/job/faculty. Inclusion—how do we get those minorities to want to be into our program/job/faculty and feel happy being there? Equity: ….? Is it the Bowtied mantra: “Equal Opportunity, unequal results”? Or is it “Equal outcomes”? Well, the general image used to explain equity is this one:
For equity, the goal is to get everyone to the same outcome. While this may seem extreme and unfair, it is rooted in general teaching theory and practice. “Equity” has only become an issue now that it is applied to helping historically underrepresented groups. But “equity” was a common teaching challenge decades before.
Equity is the logical outcome of solving the problem: “How do we get F students to become A students?”
This challenge has existed ever since we started scaling education. On one hand, there’s a question of metrics (are A students the best at that field, or instead best at taking whatever assessment is used for that field?). There’s also the usual opportunity/resources/etc questions (can you buy your kid an A with enough tutors/ can a kid get an A if he’s raised by deadbeat crackheads?). There’s also a question of reliability of metrics—is an A from Prof Harder the same as an A from Prof E.Z.A.?
Since educators can only control themselves, the biggest question in the field with regards to solving this problem becomes ‘how do we as educators make an impact on our students?’ One idea is that if you are a great teacher, you can get F students to become A students. This seems to be the holy grail of teaching: be such a good teacher, anyone can ace your class because you taught the material that well. Doesn’t matter if they were an F student, a C student, or an A student, they are all able to earn As after you are done teaching them.
Does this sound like equity to you?
The idea is that educators should make a lifelong impression on their students. Instead of delivering material that students can find for free on Youtube or Wikipedia, teachers are expected to impact the learning of their students. It’s not just presenting the material, it’s also identifying the barriers to learning the material, inspiring the students to transcend those barriers, and motivating students to learn the material. When this is mastered, the B and C students, at least, are predicted to perform as A students.
Equity as part of DEI applies this concept in the context of historically underrepresented groups.
The main problem with this theory is that it assumes the teacher contributes 100% to the students’ success. In reality, the teacher contributes 0-25% of the students’ success. Shrike’s experience is that A students are A students in every class, F students are F students in most classes, and teacher/teaching style or fit may bump a student a letter grade or two. Shrike suspects teachers have a larger part on the non-graded part—convincing a student to pursue a particular subject in greater detail or to change career trajectories. But equity does not concern itself with that part. The focus is learning outcomes, aka grades.
Let’s look at how faculty grade, and student distributions.
Grading Theories
For a hard class (defined as <1% of the class scoring 99+ on a 100 point scale), the traditional view is that you get a Bell curve for student grades. It’s the idea that the data are normally distributed, and you have a Gaussian distribution. While some argue this should be prescriptive (eg, curving scores so you have 10% As, 20% Bs, 50% Cs, 20% Ds, 10% Fs no matter how hard OR how easy), a normal distribution is expected if you only have one peak.
However, at some schools, there are two peaks to the student grades, giving a bimodal distribution. In this case, there is one peak similar to the normal distribution above, and a second peak of Fs. Note this bimodal distribution does not appear to be limited to hard courses. This is what Prof Harder’s class looks like.
There are many ways that faculty modify these grade distributions to suit their goals. The simplest method is to curve the course. The visual way is to decide what cut-off score you will use for As, Bs, Cs, Ds and Fs and then scale to fit. While in theory professors can curve the course to make it harder (eg set the cutoff for an A at 95 instead of 90), in practice this is seldom done. With the ability to curve, you can fix your grading distribution however you want it. If Prof Harder decides that 15% As instead of the 10% As will keep deans and other administrators off his back, he could curve the course:
You’ll note that with a hard class, you still have the problem of that bimodal distribution. You have to curve aggressively if you want to cut those numbers of Ds and Fs down, which means B and C students become A and B students thanks to that tail and your strong belief that Cs should be the most common grade in the course. You may note from the grade distribution above that Prof Harder is willing to accept more Ds and Fs. This is because he has an explanation for that second peak in the data (purple line), which contradicts the foundational ‘everyone can get an A’ belief of equity.
However, that bimodal distribution makes hard classes look impossible (see the black line), which upsets administrators, and everyone else who believes the metric of a good teacher is how many As they hand out.
Since some professors have figured out that students like easy classes, and happy students = good teaching evaluations, they’ve found a way to solve this ‘good teachers hand out lots of As’: They make the class easy. Now instead of a Bell curve, you flatten the distribution on the A-side. This next curve shows what Prof E.Z.A.’s course looks like. The chart is normalized to help compare to Prof Harder, but note that the scale has changed!
That bimodal distribution still upsets Prof E.Z.A because his class isn’t 80% As and Bs. Worse, Prof E.Z.A. believes that being a good teacher means transforming C-F students into A students. Since he is failing to do that, he’s tearing his hair out over what a horrible teacher he is. He’s at a loss, because he has no idea how to help those students.
Prof E.Z.A is not cynically dumbing his class down like some of his colleagues because it’s less grading, better student evaluations, and fewer student complaints. He believes in his teaching method. He gets into frequent arguments with Prof Harder about grading theories and expectations for students. Prof E.Z.A. argues that he is setting realistic expectations for the students, and teaching the material well enough that they can all master the material. If Prof E.Z.A. were teaching a foreign language, he would argue that the grading reflects competency in skills taught in the class, which is why it is possible for the entire class to get As. He’s old, so he doesn’t quite understand the newer ideas of equity, but he has been practicing equity for his entire teaching career.
This means that teaching difficulty is tied to equity. One traditional way to manage equal outcomes is to limit the amount of knowledge taught, and go over it until the most [redacted] student learns it. While there are different teaching gimmicks used, this boils down to: spend more time per “unit” of material. Now every midwit who puts some effort in can get an A.
This approach is celebrated for teaching at scale because it provides a system by which you can get most everyone to learn and retain the knowledge (represented by getting an A). And since you can demonstrate what they learned, you can push back against the ‘students didn’t earn the As’. This style of teaching is aimed at the midwits, who comprise the bulk of the students.
With the increase in that bimodal distribution, the solution is to dumb down the material even more. Possibly add a dose of nagging. However. The system will auto-adjust down until students who put zero work in can still pass the class (see that purple line on Prof E.Z.A.’s distribution). Now you understand why this well-meaning solution, combined with scale, has destroyed the US educational system. It’s worse prior to college, since some schools do not hold failing students back.
Before we look into that second peak in the bimodal distribution, let’s recap equity-based teaching:
Pros
Students retain a larger percentage of the material taught
Tailored to the middle of the Bell Curve, so works at scale
More students get As = happier and greater belief they can succeed in that field
Happier administrators
Teachers feel good about themselves and their teaching prowess
Cons
Less material taught
Bores the living daylights out of smart students
Lazy students can arb the system down to getting As for breathing
Whiny students scream bloody murder if you accidentally move them out of their comfort zone
Bimodal Grade Distribution Woes
That second peak in the bimodal distribution is an existential threat to non-cynical equity approaches to teaching. No matter how easy the course is made, there will be students who do not do the bare minimum. This will reduce the amount of content over time. It may take 50 years, which is what we have observed with the US educational system. The difficulty in math problems in the 1970s are higher than the problems now. This is the prime weakness of equity, and scale.
This means that even equity-based approaches need boundaries. These boundaries will be assaulted by students and others, often on equity grounds, which becomes a challenge for these approaches. Historical data suggests these boundaries, no matter how firm, will erode over time.
Now let’s look back to Dr Harder’s approach. Recall that the second bimodal peak is disappointing to him, but he doesn’t lose sleep over it. That is because his hypothesis is that the students in that second peak do not put in sufficient effort, aka they barely attend class. There are many ways to measure effort in the classroom today, with “clickers” being the easiest and most widespread. These also automate attendance taking. When Dr Harder maps students by effort, he finds that the second peak (purple line in the earlier graphs) is composed of students not putting in the effort. Remember, this isn’t even ‘how much are students studying?’, this is ‘do students show up to class and pay enough attention to answer questions in class?’.
Why should students who do not put in the effort receive a passing grade, let alone an A?
This is the fundamental flaw to equity-based approaches. The student contribution is lost due to the focus on the teacher’s performance. This causes a ‘race to the bottom’, and steady erosion of standards over time. When combined with the idea that some groups are marginalized or disadvantaged, it becomes heresy to equity-minded faculty to suggest the students contribute anything to their own success.
Another key flaw to equity-based approaches is that the job market is not equity-based. The job market is hyper-competitive. When everyone can get an A, you need to do more to stand out. What that ‘something more’ will be depends on the field. For biomedical science, it’s undergraduate research, and knowing how to think/design experiments.
One alternative approach to equity-based teaching is merit-based grading. Shrike will define ‘merit-based grading’ here as a grading system that separates all students by aptitude and mastery of the subject. That means the course can’t be easy, because it doesn’t separate the students (in practice everyone on Prof E.Z.A.’s curve >100 would be capped ~100). Easy courses that make it straightforward for students to get As fall under the equity bucket, even if the students need to put some effort into the course. The key for merit-based approaches is that all the students are challenged, and the grading scale separates them out. This can be curved later as desired.
The pros to this approach are that you keep the dedicated students engaged, you can cover more material, you help the students become more competitive, and students earn status by achieving high grades in your course. You also disqualify some of the half-hearted students who are not interested in your field, along with the whiniest students. The cons are that the middle students struggle to keep up, smart students who don’t want to do the work will give up and/or drop, and the course will have narrower appeal.
To summarize the Pros and Cons to merit-based grading:
Pros
Engage the smart students
Teach more material
Make the students more competitive
Give students status when they complete your course
Whiners avoid your class
Cons
Narrower appeal = lower enrollment
Students retain a lower percentage of the material
Need to aggressively manage expectations
Average students struggle more
Lose peripherally interested students
But what about the low end of that bimodal distribution? This method doesn’t help them either!
Correct.
In Shrike’s opinion, the best way to help them is to get them to drop your course and/or leave college. Since disqualifying students who are a poor fit for college drives equity-minded faculty crazy, the alternative for those faculty is to get non-performing students into a remedial course that brainwashes them with basic college survival skills. Another option is to help your department more aggressively manage these students. Common solutions include screening prior to admitting students to your major (i.e. let another program fail them out) and/or requiring Cs in pre-requisites and major courses. Another is to triage these students into a remedial course in your field if they struggle with the intro courses. Catching these problems early is key. By the time they’re juniors and seniors, it’s too late for those who never show up. Instead of wasting the time on students who don’t care, invest that time in the students who do.
Conflicts in Teaching Approach
The debate between merit and equity approaches is a staple of education because they represent two different approaches to education that have existed for decades. Both have their merits, though equity-based approaches are easier to scale.
Based on these prior differences, it is not surprising that ‘equity’ was elevated to part of the DEI trinity. Since equity approaches scale better, they will be more widespread. As they become more widespread, this becomes the gold standard for education. When educators applied their learning and insights to DEI, equity approaches were bound to be added. Beneath the rabid focus on underrepresented minorities earning As at all costs, most equity-minded faculty worry about all their students earning As at all costs. This is not an approach that will leave academia anytime soon.
And for the midwits (aka most college students), equity approaches are better. If it’s a class you don’t care about, equity-based approaches are also great. These approaches benefit a greater number of students than merit-based training.
But if you’re trying to be elite, you need merit-based courses in your field. You need that competitive edge that will help you beat your peers who are as smart, or smarter than you. You need to learn more than what the midwits are expected to handle. If you’re trying to be elite, your competition is doing the same thing. It’s a race, and you want the courses and teaching styles that let you master as much material—in the lab and in the classroom—as you can.