Eyes Wide Shut

How Anti-racist Intentions Can Increase Structural Racism

Preludes and Warnings

Conjecture: Many policy decisions which are implemented with the intention of advancing anti-racism are instead supporting structural racism because proponents have failed to think critically about terms, definitions, and causal structures. Moreover, the magnitude of this problem is proportional to the probabilistic/statistical reasoning needed to articulate the issue under consideration.

Now, I have no hope of "proving" this conjecture here, but what I will do is provide a relevant data point which is consistent with the above hypothesis, and which is a serious problem on my campus. That is, senior administrators (who desperately want to be seen as anti-racist) are measurably contributing to structural racism, and this is largely due to thinking sloppily and dogmatically about an issue which requires more critical thinking and less critical theory. Specifically I'll discuss the cascade of problems generated by allowing the university math placement exam to be taken without proctoring.

I'm going to acknowledge up front that the majority of potential readers will find this to be a dull topic. So it goes. My goal is to try to be as accurate as possible in identifying the problem while still conveying the non-trivial damage it's doing to education at my institution. My intent is not to generate ammunition for the academic front of the modern culture wars. I want to show you a very real problem, and spelling out the details is a rather dry process. Skim if you want, or if that's not really your bag then move along to something flashier. For those readers that remain: here we go.

The Causal Problem

The primary focus of this post will be on the structure (and consequences) of the university math placement exam. What is it? Generally speaking, incoming first year university students will need to take some sort of math class (especially if they aspire to major in STEM), but they need to figure out which course to take first (for example, many need to be placed somewhere on the track of algebra -> pre-calculus -> calculus). There are a few mechanisms for this sorting, but the standard one is to take the university math placement exam. That is, take the exam, get a score, the score tells you which math class(es) you are prepared for, and students register accordingly. Seems reasonable enough.

Except.

Except my institution allows students to take the placement exam unproctored; that is, without proctoring. Repeatedly. At home. And up until very recently, untimed. In other words, university policy was basically "Go to this website, take this online test at home, as many times as you like, without anyone watching, and use as much time as you want. Your best score will place you into the appropriate math course." Actually, it's worse than this. The Student Success page (for aspiring STEM majors) says:

1. "It is highly encouraged that science students try to begin at the [pre-calculus] level or higher.”

2. "Students who place beneath the math level they expected should complete the suggested Learning Module on the ALEKS website and retake the test."

That is, my institution is encouraging prospective science students to start in pre-calculus or above, and to repeatedly take the unproctored at-home math placement exam until they get a score they are happy with. (You can verify this language here.) When I discovered that this was policy I was kind of floored, but one good thing was that my institution actually collected pretty detailed data on the exam. For example, I could tell that there was at least one student who took the exam eleven times, and another that spent twenty-four hours on what should have been a ninety minute test. Not exactly stringent guidelines.

Anyway, it doesn't take a genius to figure out what happened. See, as a general rule, U.S. students tend to dislike math, and if one gets placed into a low level math course they will have to spend more time (and hence more money) to complete their full math requirements. As such, some combination of financial incentives and motivational hedonism implicitly encouraged students to cheat on the placement exam. And the reality is that there were absolutely zero checks in place to tell if Jonny took the placement exam honestly (the way it was intended), or if he googled the answers, or used Wolfram Alpha to assist, or if Jonny just had his cousin take it for him. Gee... What could possibly go wrong?

I should pause for a second and note that perhaps "cheat" is too strong a word. It is and it isn't, but I should attempt to put a human face on this because it highlights how easily it is for the placement test to lack validity. See, I think some students took algebra (and perhaps even pre-calculus) in high school and expected to place into calculus in college. But when faced with the placement exam, they discovered that there were questions that seemed familiar, and they knew they should know how to solve them, but they couldn't. And what they needed was a little "reminder." You know... A little help. Enter Google, or Chegg, or an old textbook or whatever. My point though is that some students might have honestly thought that getting "hints" or "reminders" from the web during the exam (which would be at least functionally equivalent to cheating, if not ethically equivalent) were resulting in a exam score which better reflected their mastery level. I think it's quite possible students thought this way.

Except of course that such web-assistance absolutely does not yield a more accurate placement score, and so these students were setting themselves up for failure. Or rather, the university was letting students set themselves up for failure. A reasonable skeptic might hear this and respond "Well, that's a nice hypothesis; it sounds plausible, but is it actually happening? Do you actually know?" Good questions. I'll respond.

Discovery of the Causal Problem

As a general rule, administrators are pretty disconnected from actual instruction (at least in mathematics), and so they mostly care about really broad metrics of institutional success, like the so-called "DFW rate". This is the percentage of students who enroll in a course and who end up receiving a final grade of D or F, or withdraw from the course. It's a pretty crude measure of student success, but also a fairly common measure from an institutional perspective. Importantly, the DFW rate in lower division math courses (and many STEM intro courses) at my institution were incredibly high. I mean really really high. I'm told by knowledgeable colleagues that our DFW rates are actually fairly in line with comparable institutions, but I think that both the public and our senior administrators (and I) think our rates are simply off-the-charts too high. In any case, a few years back the topic of sky-high DFW rates was rampant on campus and so the Math Department (and perhaps other departments as well) came under significant scrutiny to identify the cause of these high rates and take steps to reduce them.

Not surprisingly, there seemed to be a number of contributing issues, and they all were either created or exacerbated by administrative failures. Sometimes it was Math Department administration that failed, and we will own those failures and work to the best of our ability to correct them. That said, it's also worth mentioning that my institution is a research university and yet it went twenty-five years without hiring a single research mathematician. So let's be honest: you get what you pay for, and my university essentially paid nothing. Maybe I'm exaggerating a bit, but I'm not exaggerating in saying that up until just two years ago, the tap water in the Math Department's building was not potable. There were signs above all the sinks and drinking fountains in our building saying "Don't drink this water." I am not joking.

In any case, the DFW rates were super high, so the Math Department researched the problem. We developed our own standardized "quick" assessment diagnostic exam that we administered to our students within the first week of class. We discovered that our custom tests were predictive of student success (for example in grades on the first midterm) but that the unproctored ALEKS exam (which is the university math placement test) did not. Or rather, the best score on students' ALEKS exam was not predictive of future success (basically zero correlation). In contrast, students' first ALEKS exam score was predictive of course success. So you can see what was likely happening: Students would take the ALEKS exam, sometimes getting a bad score which suggested they should not be taking advanced courses, then they would take the exam again -- miraculously getting a much better exam score that placed them into higher courses -- and then suddenly their performance in that class was not as strong as their ALEKS exam score predicted. In fact, their performance was much worse than predicted. Just like their first ALEKS exam score suggested.

We also started just asking our students if they had "help" on their ALEKS placement exam. For example, after a month or two into a course, some students would realize they were severely struggling. They'd ask me "Do you think I can pass this class, or do you think I should withdraw?" This usually initiated a longer discussion about study habits and preparation and such, and I would try to casually ask about the manner in which they took the ALEKS test. Every time I asked, the students who spoke to me admitted to using some sort of external help on that placement exam, which again indicates widespread cheating.

Finally, we had an enormous amount of anecdotal evidence from math faculty pointing out that many students seemed to not have the prerequisite knowledge to succeed in their courses, and this too was reflected in our diagnostic exams. For example, we had a shockingly high percentage of first-semester calculus students who could not solve a linear equation in one variable. Seriously. In short, the math placement system we were using was broken.

A Proposed Solution and What Happened Next

The math department wrote up our findings (including a fair amount of pretty basic statistical analysis) and reported them to our administration. Our primary recommendation for reducing DFW rates was to require proctoring for the ALEKS placement exam. No more letting Jonny's cousin take the exam for Jonny. The math department approved. So did the dean and senate of the College of Science and Mathematics (CSM). The provost's office was not thrilled, and they balked. We entered into discussions with the provost's office, and after months of dialogue, we finally convinced them and the provost agreed to require proctoring for the test. Good news, right?

Well, it would have been, but shortly after we finally got approval to require proctoring for the placement exams, we had a regime change. A new provost came in, and he reversed the previous provost's decision before it could be implemented. Proctoring would not be required.

Now, there are two main threads which stem from this decision. The first is how much damage this does to our student population, and the second is the provost's incredibly poor reasoning to justify it. I'll start with the damage.

Downstream Damage From Lack of Proctoring

It's important to remember that the initial discovery of the problem extended from administrative pressure to resolve sky-high high DFW rates in the Math Department. Administrators tend to forget that high DFW rates are not a problem in and of themselves, but rather they are a symptom of a problem. In this particular case, the real problem was a broken math placement system, which was dysfunctional because the placement exam was unproctored and a non-trivial number of students were effectively cheating on it. Importantly, once the math department identified this problem, we could then look at all the downstream consequences that extended from it. We'll discuss these now.

To gain a mental picture of these consequences, let's consider the extreme (but not exactly rare) situation in which a student really should be placed into college algebra but instead is placed into calculus. What is the course like for that misplaced student? To get an idea, one first needs to understand that the mathematics needed for STEM majors follows a linear progression: arithmetic -> algebra -> trigonometry and pre-calculus -> calculus I -> calculus II -> calculus III -> differential equations -> and so on. Importantly, mastery of knowledge in successive courses requires mastery of knowledge in prerequisite courses. It's kind of like constructing a multi-level building: You can't start building the fourth floor if construction of the third floor hasn't even started. And this is really the core problem: Letting students cheat on the math placement exam sets them up for failure by officially placing them into courses they cannot pass. The university provides a gold stamp of approval that says they are ready for calculus when in reality they need to retake algebra. They will take that calculus class, and in order to be successful they will need to learn not just the new material of calculus, but also trig and pre-calculus, and also college algebra. And because of the linear structure to this material, they cannot learn these subjects concurrently. For example, algebra is simply the bread and butter of calculus, and students will simply not be learning calculus at all if they do not have the requisite algebra skills. Guaranteed. So what happens to this misplaced student? They need to learn algebra but their instructor is teaching calculus so they will be learning nothing.

So they fail — but it's so much worse than just failing.

See, they need to learn algebra to succeed, but they aren't learning it, so they fail, but they think they should be enrolled in calculus, where they will fail and where they will not learn what they need to succeed. So after they fail, they retake a course they are bound to fail again. And again they aren't learning what they need to succeed, so they retake calculus just to fail it again. Typically they repeat this until they drop out (sometimes of college and sometimes of just their major) or else an instructor gives them a passing grade they haven't earned which thereby creates the same problem in a more difficult STEM class.

In short, my institution has set this student up for perpetual failure. Moreover, my institution is doing this knowingly. The reality is that sometimes institutions have policies which 1) hurt student success, and 2) the institution is largely unaware that such policies are doing so. But that's not the case here. In this case, the math department has identified the problem (including providing a statistical analysis to support its assessment), and it has gone through all the appropriate governance procedures to fix the problem and everyone (math department, College of Science and Mathematics, and the previous provost) support the proposed solution of proctoring the math placement exams. But our current provost purposefully reversed that decision, and thus he is personally responsible for the damage this causes to student success. That is, our current provost has decided — against the recommendations of every other informed decision maker in the chain of governance — to set up many STEM students for perpetual failure.

It gets worse though.

See, the provost has decided to place these students into a perpetual DFW cycle in which students repeatedly fail a course they cannot pass, but my institution also keeps collecting tuition dollars in the meantime. That is, my institution has a mechanism to place students into classes they cannot pass and take their hard earned tuition money in the process. It's unethical, and it's a social injustice. Moreover, it is structurally racist. How? Well, it's important to understand that my institution is majority-minority. Because the population of the U.S. is not minority-majority, we see that any damage done to our students' learning has disproportionate impact on minority students. Institutional policies which disproportionately hurt minority students are structurally racist by any definition of structural racism I've seen. This is inescapable.

When possible, I like to use actual quantities to really ground the discussion, so let's provide an estimate of the magnitude of this social injustice. First, it should not be surprising to discover that nearly every major in the College of Science and Mathematics requires at least a semester of calculus. What might be surprising is that that same college is the most racially/ethnically diverse college on campus. More specifically, of the students in the College of Science and Mathematics, only 28% are white, and of the non-Asian students in this college only 35% of the students are white. That is, by whatever definition of "student of color" you might choose, students of color comprise a definitive majority in the college which (almost uniformly) requires calculus. And setting up students for failure in calculus (as our provost has decided to) thus overwhelmingly affects students of color, and is therefore definitively structurally racist. There is simply no way around this. And remember: the current provost made this decision against the recommendations of the math department, against the recommendations of the college senate and dean, and against the decision of the previous provost. Our current provost owns the consequences of his decision, and those consequences are an increase to structural racism.

Percentages are nice, but let's try to put some dollar amounts on the damage being done by the provost's decision. Just to give an initial ballpark estimate, keep in mind that our annual in-state tuition cost is about $15,000, and a full-time student takes about 15 credits per semester. That works out to be about $500 per credit, and single variable calculus is a four credit course. This means that each time a student is misplaced into calculus, the university is fleecing them out of about $2,000. We seem to be teaching about 23 sections of single variable calculus this semester. If we make a conservative estimate of only one misplaced student per section of calculus I, then we see the provost has decided that it's okay to needlessly take about $46,000 from students this semester alone. That's pretty close to $100,000 per year for calculus I, and the problem extends much farther beyond calculus I.

Indeed, to get a better understanding of the scope of the problem, we have to look at what happens to students who are misplaced by our system. As I've said above, they essentially repeatedly take the math course (say, calculus I), and typically one of two things happens. Either eventually they realize that they will never pass this class and then drop out of college or (sometimes) change majors, or else eventually some instructor takes pity on them and lets them pass the class even though they don't know the material. But keep in mind that passing a student in calculus I when they don't know the material (particularly when they don't have a firm grasp of algebra) just creates a problem for any course which requires calculus knowledge in order to be successful. And importantly, STEM majors tend to require knowledge of calculus in order to ultimately be successful. This means that the problem of misplaced-and-not-prepared-student will occur throughout STEM majors. That is, what at first seemed like a math-department-only problem is suddenly identified as a serious problem for all STEM disciplines. The decision to let students cheat on the math placement exam then effectively allows the university to fleece tuition money from STEM students by setting them up for failure in math, physics, chemistry, computer science, engineering, and biology. Remember that ~$100,000 fleecing per year was a conservative estimate for one course (Calculus I) in one discipline (math) over one year; extended out across all relevant courses in all relevant disciplines and that estimate starts to look like about a million dollar take every year. That’s a fleecing which is opposed by the Math Department, opposed by the College of Science and Mathematics, and opposed by our previous provost, but which is supported and enforced by our current provost.

Although I strongly believe that this is the most grotesquely unethical aspect of the provost's decision, I think it's important to point out that the damage extends further. See, when you set up students for failure, they tend to fail, which in turn raises DFW rates. Remember, administrators tend to think that the DFW rate is the problem (presumably because it makes it difficult for the administrator to get their next job), and thus they put pressure on faculty to lower their respective DFW rates. But when the DFW rate is too high because students are being misplaced into courses, that administrative pressure can only be relieved in a few ways: Water down the curriculum, lower standards, or pass students who simply don't understand the material. Each of these pathways fundamentally undermines both the quality of the education we provide and the value of the degrees we award. Indeed, a degree in a STEM field should indicate that the student with the degree has actually learned the relevant material for that major, but if administrators demand a decrease in DFW rates while placing students into courses they can't pass, then the only possibility is the awarding of a degree which has no value. And that's the direction my institution is headed. Indeed, attend few Board of Trustees meetings, and you will see what the high-level administrators care about: Lower DFW rates, higher retention, and increased graduation rates. What they don't seem to care about is maintaining standards of excellence and whether or not students are actually learning the material the public has entrusted our institution to provide. Put simply: Higher education is more about credentialism than actual education. "Institutions of Higher Credentialism" sounds about right.

A final downstream problem with allowing students to cheat on the math placement exam is that it fosters a culture of dishonesty and unethical behavior. The problem is that if the university tolerates cheating on an important placement exam, then the students who cheated learn that cheating is accepted on campus, and thus they will cheat again on later course exams. Students with integrity will then discover that 1) they are competing against students who are cheating, and 2) the institution allows this dishonest behavior. Students with integrity will slowly learn that cheating is acceptable, and so they too will cheat. This is a disaster for maintaining academic integrity on campus. Unfortunately this accidental-promotion-of-cheating was recently compounded by our provost's decision to roll out a new "restorative justice" approach to violations of academic integrity which made it exceedingly easy for habitual cheaters to game and allowed administrators to overturn faculty sanctions on cheating students. This was all part of a commitment to anti-racism, and it fairly obviously did more damage to our institution than good. And the damage is palpable; just ask faculty if students have been behaving more unethically in the past couple years and the answer is almost always yes.

The Provost's Response and Rationale

Before discussing how the provost has justified his decision, it might be helpful to first briefly discuss the power structure of the university. Every university has the equivalent of a "Chief Executive Officer" who is essentially in charge of the entire campus. At our university this role comes with the title Chancellor. (We are also part of a public university system, so there is also someone who oversees all the campuses in our system, and this person has the title System President.) On each campus there is also a person with the role of "Chief Academic Officer" who is in charge of essentially everything academic related; this person has the title Provost and reports to the chancellor. Now, the rather interesting thing is that on the academic side of the house, the provost has an incredible amount of power. Essentially the provost has ultimate authority in decision making as it pertains to academic issues. Of course it would be impossible for the provost to make every single academic decision on campus, and so the provost has a whole office of vice/associate/assistant/junior/quasi-provosts, and many decisions are delegated downwards to deans of colleges and then further delegated downward to department chairs. It's also worth noting that attached to various levels of governance are deliberative bodies (comprised of faculty) with the purpose of providing governance recommendations to university administration.

You may have heard of terms like "shared governance" and "departmental autonomy". Basically the idea is that there are a variety of decisions for which departmental faculty are best suited for making, and the provost should generally respect those decisions (that is, not overturn them when he disagrees). Unfortunately, this doesn't always happen and the provost will make decisions which go against faculty recommendations (even when it's the faculty who have the best insight and expertise on a given issue). What faculty often don't understand is that the provost is completely within his (or her) power to ignore any of these recommendations or lower level decisions — and there is essentially no appeal. In short, on the academic side of the house the provost is basically god. Okay, "god" is too strong, but "Caesar" gets pretty close to it. Consequently, this gives faculty (who can see the damage being done) very little power to influence the provost and convince him to change his mind. Indeed, levers of power are scarce, and they mostly consist of 1) coercion from accreditors, 2) frustration from higher up the political food chain, and 3) general public anger. Perhaps it then becomes more clear why I'm writing such a detailed public post on the matter.

In any case, all this discussion of power structures on campus is to highlight the fact that the provost simply has the power to make such decisions — even when they are contrary to the desire of every other governance body on campus — and technically he doesn't need to provide any justification. Essentially, his will be done. (Side note: Actually, what I've said is not quite true, in that when the provost overturns a decision (or some other official recommendation provided by a lower governance level) our founding governance documents require him to provide written justification. I suspect this documentation requirement is in place in part to purposely burden a provost who is (at risk of) abusing their power and to also provide a clear record of such reversals. Unfortunately, most faculty on my campus seem to be unaware of this documentation requirement and so the provost simply ignores it. The result is a de facto Academic Caesar.)

Still, one wonders why the provost has decided to to allow students to cheat on the math placement exam when it's unethical and arguably theft from a minority-majority population. The off-the-record justification seems to be that the administration thinks "standardized tests are racist". I haven't personally heard this argument voiced (which is probably a good thing for my blood pressure), but my colleagues have. It's also worth noting that such reasoning seems inline with a recent Strategic Plan For Racial Equity from our State Department of Higher Education which explicitly calls for the elimination of all standardized tests (for admissions, assessment, and placement) in college-level math and English. (See page 41 of this document.)

Just to be clear: This decision by the Department of Higher Education and by the provost is unequivocally anti-scientific. There is no way that the consensus opinion of psychometricians (the scientists who research standardized test reliability, validity, measure invariance, etc) would support the elimination of standardized tests for admissions, assessment, and placement for mathematics, nor would they support a blanket statement like "standardized tests are racist". I am entirely certain that the administrators, faculty, and stakeholders who helped craft this state level decision have the absolute best intentions for racial equity, but because they simply don't understand mathematics, statistics, and psychometrics, they are making a political decision which runs completely counter to consensus science, and it will measurably contribute to structural racism. I am absolutely beside myself that the state that leads the country in education efforts could make such a terribly damaging decision, but I guess that's the nature of the beast today. Measurable damage to marginalized communities doesn't matter if the anti-racist performance it's wrapped in is sufficiently convincing. Remind me again if it's impact or intentions that matter more.

The provost's on-the-record justifications for not requiring proctoring tend to sound a bit stronger — at least on first pass — but they don't stand up to scrutiny either. For example, the first (and likely dominant) reason the provost has provided is that he simply doesn't believe students are actually cheating on the placement exam. I suppose that's the fastest way to dismiss the proposal to proctor exams: Just deny that a problem exists. Of course, the Math Department has sent multiple reports on the matter including a statistical analysis identifying the problem, and every math faculty member can point to myriad examples of students in courses being completely unprepared. But none of that evidence seems to matter. I suppose this is what he means when he says his decisions are "data informed and not data driven" — inconvenient data is just ignored.

It looks rather bad to tell faculty (specifically those faculty who have provided substantial evidence of the existence of a problem) that the problem doesn't actually exist, so the provost has tried a few other justifications as well. For example, he once tried to argue that the research supported his decision. I wonder if he forgot that we are faculty in the College of SCIENCE and Mathematics, because we then asked to see the supporting articles so we could read them. He sent them, we read them, and NONE of it supported his decision; some of it actually undermined his decision. I ask you: What message does it send to scientists when an administrator misrepresents empirical research to support their decision?

Another of the provost's attempts at justification has been to cite "disparate impact". This tack can catch faculty off-guard if they haven't heard of it before because it is a doctrine which makes it illegal to use tests for employment/admissions/placement if doing so would disproportionately reduce the percentages of students in protected classes. Moreover, it's also true that historically marginalized students on average tend to score lower on such academic tests of mathematical knowledge (such as math placement tests, the Math SAT, etc) than white and Asian students. The combination of these two facts then often generates an uncritical gut response to avoid the use of standardized math (placement) tests, or at least to rely on their use much less and certainly not more. Unfortunately, such a response would be a mistake. Indeed, the disparate impact doctrine has some important caveats; for example, tests which disproportionately effect some protected classes can still be used if the test is shown to be a valid predictor of relevant future performance. So sure, you can't use Math SAT scores to decide who gets on the football team, but it's nonsensical to believe that the disparate impact doctrine prevents a valid math placement test from being used to place students into university math courses.

As I'm writing this, a disturbing thought occurs to me. The Math Department discovered that because students are cheating on the math placement test, the current version of the administration of the test has lost its predictive validity. But if it's also the case that on average historically marginalized students are performing worse on the test than white (and Asian) students, then the disparate impact doctrine would make it illegal to continue using unproctored ALEKS. On one hand this would be a legal argument for requiring proctoring for the ALEKS placement test, but on the (disturbing) other hand it could be used by anti-standardized-test ideologues to eliminate the use of ALEKS completely. I don't like the thought of this, and I don't like making the suggestion, but I also can't (yet) rule it out as a reason why the provost wants to prohibit proctoring. Still, giving the provost the benefit of the doubt, it would then seem that there is even a legal case against the provost's decision to not require proctoring.

Recently, the provost has been testing out another justification for not requiring proctoring for math placement, and it also sounds good if you don't think about it too much: There's not enough money. Now, it's true that our institution is strapped for cash, but there's an easy way to implement proctoring which is revenue neutral; namely, pass the cost on to the students. Now, you might think that it's completely unfair to place this additional burden on students, but it's important to realize that even if we were to do this, the net cost to all students would actually decrease. Let me explain.

Suppose we charge each student $15 to take a ninety-minute proctored math placement exam. Suppose that the Campus Testing Center hired someone to proctor the test for four students at a time. That's $60 for ninety minutes of the proctor's time; or about $40 per hour (a pretty reasonable wage). Now $15 per student is not nothing, right? True, BUT each time the proctored placement exam directs a student to the proper initial math class when they would have otherwise been set up for failure (due to non-proctoring) we are saving that student about $2,000 in tuition (for example, if they would have gone into a four credit calculus course but were only prepared for College Algebra). That means each correctly placed student (who would have otherwise been placed incorrectly) saves enough money to cover the proctoring costs of 133 other students. We can phrase this another way. Suppose all 30 students in a calculus I class were required to take a $15 placement exam. That amounts to an additional cost of $450 to those 30 students (because 30 times $15 is $450). However, our conservative estimate was that one student in each calculus I class had been been misplaced into that class and a proctored placement exam would instead correctly place them into a lower level course. Correctly placing that student then saves that student (and hence the batch of 30 students under consideration) $2000 in tuition. Thus the net savings for these 30 students is $1,550 = $2,000 - $450. And this is just one section of one course in one subject area. And again, proctoring would lower the DFW rate, increase retention rates, and increase graduation rates. In other words it actually makes financial sense to require proctoring for the math placement exam.

There's a final point I should make about the existence of limitations on the provost, and the point is that some limitation do exist and are generally enforced by accreditation standards. That is, the provost is essentially the Caesar of Academic Affairs on Campus, but he cannot run afoul of accreditation standards. Except that by letting students cheat on the placement exam he has indeed run afoul of them. My institution gets its accreditation from the New England Commission of Higher Education (NECHE), and they post their standards for accreditation publicly. (See here.) I won't bore you with too many details, but it seems clear that the provost's actions are in violation NECHE's standards. For example, Standard 4.44 contains the language:

"The institution works to prevent cheating and plagiarism as well as to deal forthrightly with any instances in which they occur. It works systematically to ensure an environment supportive of academic integrity."

Our provost is encouraging the exact opposite of this by ignoring the systematic cheating that is occurring on campus, allowing it continue, and taking steps to decrease academic integrity. The introduction to Standard 5 contains the language:

"[The institution] endeavors to ensure the success of its students, offering the resources and services that provide them the opportunity to achieve the goals of their educational program as specified in institutional publications."

But instead of endeavoring to ensure the success of students, the provost has decided to set up many of them for failure by allowing them to cheat and place into courses they will never pass. The same Standard also states:

"The institution’s interactions with students and prospective students are characterized by integrity and equity."

But again, because this math placement issue disproportionately damages both the education and finances of marginalized students, the provost's decision is inherently inequitable.

In Summary

Alright, let's recap:

1. The Math Department discovered that because the university math placement test has not been proctored, many students have been cheating on the exam and were being placed into courses they could not pass.

2. When students are misplaced into courses they cannot pass, they will repeatedly fail or withdraw until they drop out or are passed without having learned the material (thereby reproducing the same problem of perpetual-failure in a subsequent course).

3. Students pay full tuition for courses they are set up to fail in.

4. Because the student population in the College of Science and Mathematics is 72% non-white, the process of charging students tuition for courses they cannot pass disproportionately impacts historically marginalized students.

5. Allowing students to cheat on standardized exams fosters a campus culture of dishonesty and undermines academic integrity.

6. Setting students up for failure in math courses increases DFW rates, lowers retention and graduation rates, reduces student success, and puts pressure on faculty to lower standards, inflate grades, or simply pass students even though they haven't learned the material.

7. The Math Department identified this problem, provided a report on it including statistical justification, and proposed that the placement exam be proctored so as to resolve the issue.

8. The College of Science and Mathematics (senate and dean) approved of this proposal.

9. The previous provost approved it as well.

10. The current provost disagreed and has refused to require proctoring, thereby deciding to perpetuate 1) academic dishonesty, 2) increased DFW rates, 3) decreased retention rates. This decision violates multiple NECHE accreditation standards, and it is arguably theft of student tuition.

11. The provost has provided a variety of attempted justifications for his decision, but none of them stand up to scrutiny. Careful consideration of them suggests that there are good ethical, financial, and legal reasons for the provost to reverse his decision by requiring proctoring.

It should be clear that allowing students to cheat on the math placement exam causes a cascading disaster for students, faculty, and administrators on our campus, and the effects are particularly strong in STEM disciplines. Moreover, because our institution is minority-majority, the damage wrought by the provost's decision will disproportionately land on historically marginalized students. The dark irony here is that for the past two years our current provost has been non-stop messaging on how he wants our university to become a national leader in anti-racism. Unfortunately, I think this sort of social iatrogenics is fairly common among vociferous proponents of anti-racism. I suspect there are many reasons why this might be the case, but I think a major contributing factor has been the rise (and, in certain fields, dominance) of social scholarship which has eschewed mathematics, statistics, and quality scientific empiricism. This has completely atrophied the academy's ability to reason about complex causal networks, like the role standardized tests play in higher education (or more generally the role higher education plays in society at large). Without a careful empirical language to discuss our social reality our ability to discover truth is crippled, and instead we are left to essentially rely on academic power politics. That is, truths are no longer those beliefs that comport with reality, but instead truths are whatever leaders of one's ideological tribe say they are. A major consequence that follows is that the (perhaps large) majority of proposed anti-racist policies are in fact structurally racist because careful causal analysis has been replaced with a tribal belief which in unmoored from the reality of cause and effect.

In light of this proposed framework, I think the provost's decision makes much more sense. Whether or not outcomes for (marginalized) students improves is a secondary concern to whether or not decisions appear to be anti-racist. The empirical science of psychometrics is replaced with ideological beliefs about standardized tests being racist. Statistical analyses of the problem are rejected in favor of tribal "truth". This is how performative anti-racism (which is nevertheless structurally racist) can be preferred over actually impactful anti-racism. And of course, it is society (and in our case our student body) that suffers.

It is nearly impossible for faculty — the frontline workers in higher education — to push back against a provost who is dead set on ignoring empirical evidence that his decisions are causing more damage than good. That's just how the university is structured. But in this particular case of unproctored math placement exams, the damage is so severe and so extensive, that the correct decision can easily be discovered by anyone willing to spend a few moments to consider the downstream consequences. And that's a large part of the reason why I've tried to carefully write up this example... In the hope that someone with influence over the provost can point out how problematic his decision is. Once a person in a position of power has decided that evidence, expertise, and critical thinking will be disregarded whenever convenient, the only game left is power politics. So here we are.