Gorsuch, TransAm dissent

Attention Deficit Warning: you’re about to read a long boring text with no pictures.

In Neil Gorsuch nomination hearings, two of his opinions were seized on by his detractors. One is “TRANSAM TRUCKING, INC. v. ADMINISTRATIVE REVIEW BOARD, UNITED STATES DEPARTMENT OF LABOR“, where it was alleged that Gorsuch, feeling an instinctive sympathy for large corporations and disdain for the common man, arrived at an absurd conclusion that a driver could be rightfully fired for trying to avoid freezing to death, and in doing so disobeying instructions.

The following is an attempt to ascertain if this criticism has merit.


Alphonse Maddin, an employee of TransAm, was stuck on I-88 in Illinois, with frozen breaks on his trailer, and a non-working auxiliary power unit, which was supposed to provide heating.  He reported his situation to TransAm, which sent a repair-person to aid him.  He eventually, feeling numbness in his extremities and difficulty breathing because of the cold, became concerned about continuing to sit in an unheated vehicle, and advised his supervisor that he was leaving to seek help.  His supervisor told him not to leave the trailer and to either drag the trailer with the frozen breaks, or to continue waiting for the repair-person.  Maddin disobeyed these instructions and instead, drove the truck without the trailer in an attempt to find help.  He was fired for leaving the trailer.

Opinion and Dissent

At issue here was a decision made by an administrative law judge (ALJ) and Department of Labor Administrative Review Board (ARB), which found that the driver’s termination was wrongful, because it’s chief cause was a “protected action” (an action which could not result in termination) under the Surface Transportation Assistance Act (STAA).  The statute which was deemed to have been violated makes it “unlawful for an employer to discharge an employee who “refuses to operate a vehicle because . . . the employee has a reasonable apprehension of serious injury to the employee or the public because of the vehicle’s hazardous safety or security condition.” “.

The majority opinion found that the termination did indeed violate this statute.  To do so, it interpreted the wording “refusing to operate a vehicle” broadly, to include “refusing to operate in the prescribed manner, instead operating it in a manner that would alleviate the safety hazard”. To support this reading, the opinion cited the case “BEVERIDGE V. WASTE STREAM ENVIRONMENTAL INC.” as precedent.  This case discusses a nearly identical situation, and indeed supports the broad (and in a way counter-intuitive) reading.  From the cited opinion:

Under the ALJ’s reasoning, a refusal to drive an overweight vehicle would not be covered if the load was reduced by the employee to a legally acceptable level and then delivered. We do not agree. An employee who refuses to drive illegally does not lose his STAA protection by correcting the illegality and then proceeding to drive.

Gorsuch, in his dissent, disagrees with this broad reading. Instead, he reads the text narrowly, relying on the dictionary definition of the word “operate”. It’s hard to claim that this in itself is objectionable, given Gorsuch’s originalist/textualist philosophy, however several other aspects of his opinion seemed disconcerting.

First, the tone of his opinion is brash, condescending and irritated.  It is full of insinuations to the effect that Maddin’s claims are frivolous, and that the fact that the case bubbled up all the way up to his court, without having been resolved in favor of TransAm, is in itself a malfeasance.  From his dissent:

 “…there’s simply no law … giving employees the right to operate their vehicles in ways their employers forbid. Maybe the Department would like such a law, maybe someday Congress will adorn our federal statute books with such a law. But it isn’t there yet.”

The word “adorn” is telling.

Second, there is no mention in his dissent of the precedent cited in the majority opinion.  Nor is there an explanation of this omission.  During his confirmation hearing, Gorsuch stated that if the wording of a statute is clear, the work of a judge stops there, and he needs to look no further.  However, “clear” in this sense would suggest “clear to the judge hearing the case”, not “clear to any reasonable person”.  The fact that there is an opinion which interprets the wording in a different fashion would suggest that it is not universally clear.  This “clear to me” standard seems to allow for ignoring precedents in a whole variety of cases at the whim of the judge.  For example “‘Freedom of speech’ is clear, thus there is no need to rely on precedents establishing exemptions for hate speech, libel, obscenity…”.

Third, the opinion goes out of it’s way to insinuate that the temperature experienced by Maddin, while being below comfort level, was not life threatening, without explicitly stating so, or providing justification for this.  For instance, he describes Maddin’s option of staying in the truck as “unpleasant”.  While the supposition that the temperature wasn’t life threatening isn’t strictly necessary to support his reasoning, the fact that Gorsuch chooses to insinuate this in several instances, suggests that he thought that his reasoning would not be sufficiently convincing without it.  But in fact, Gorsuch contradicts himself.  In the following passage, criticizing the majority opinion, he at the end, grants them that Maddin’s actions were indeed prompted by safety concerns, and it would stand to reason that if remaining in the vehicle was a safety hazard, it would be because the temperature was indeed so cold as to be life (or disfigurement) threatening.

To be sure, my colleagues invoke the statute’s purposes — employee “health” and “safety” — and suggest the result they reach is consistent with them. After all, they note, the employee here who chose to defy his employer’s instructions and drive his truck as he thought best didn’t do so to write a novel or with some other esoteric end in mind, but because he bore safety concerns. Just the sort of employee safety concerns, my colleagues indicate, Congress intended to protect. 

Even supposing all this is true, though, when the statute is plain it simply isn’t our business to appeal to legislative intentions.

I assume that “Even supposing all this is true”, is shorthand for “All of this is true, as there is no evidence to the contrary”, because in the text of the case I did not see any such evidence.

In short, I found this opinion of Gorsuch really poor, and inconsistent with the majority of the praise he received at the nomination hearing. It’s possible that he wrote it on a bad day, but he stood behind it without any qualms at the nomination hearing.

Posted in law, politics, Uncategorized | 2 Comments

How people think about scientific concepts

A couple of times a year, I taught a class in introductory physics for non-science majors. Once for this class, I was discussing a problem in Newtonian mechanics with some students.

I think maybe we were looking at a moving cart that suddenly had an unbalanced force applied to it. Maybe I asked what the effect of having this unbalanced force be large rather than small would be. Jason* said “the cart would move faster”

And I said something along the lines of “I see where you’re going here, but in this case we want to be very precise in that what we’re talking about isn’t the speed, it’s the acceleration

At which point Brock* turned to his neighbour, and said in mock outrage: “Dammit, Jason*, words have meanings!”

I hope you can see that this was really funny.

But it also underscores the way most people think about scientific concepts. Whether because I’ve been studying physics for long enough that I don’t notice it any more, or because English is not my native language — I don’t know why — but it doesn’t bother me that velocity and acceleration mean totally different things and you can’t use one when you mean the other. It sure does bother a lot of other people, though.

I thought about the episode the other day when I overheard two administrators talking in the campus coffeeshop. They were (I think) discussing an administration effectiveness seminar that they had just been at. Janet* said “what are we supposed to do about learning objectives*?” To which Cindy* replied “no, we shouldn’t use learning objectives at all, we need to use learning outcomes*!”. To me this conversation sounded funny. Not only was it funny, but it made me want to turn to them and say “Dammit, Janet*, words have meanings!”

The precise meanings of scientific terms seem just as arbitrary to people. “Force” and “power” are both strength words. Trying to say that they are different things entirely, and to say that you are wrong when you say power when you mean force comes off as ridiculous pedantry.

*Names changed to protect the innocent. Terms changed cause I forgot the original ones.

Posted in INMJTEY, physics | Leave a comment


Grades are a bad system. Before I began teaching, and even for a while after, I considered being against grades to be the philosophical realm of kumbaya-singing hippies (by the way, why is that song such a strong cliché associated with a particular mindset?). Something like “why can’t we all get along” and “minds are like parachutes”. But I’ve only taught for a couple of years, and now I’m somewhat against grades as well. The problem isn’t that evaluation of students is a bad idea. Evaluation of students is a great idea. It’s that the current system of grades is not a very good way of doing it.

The goal of a grade system as I see it from the student’s perspective is to allow the student to signal their ability to others. And that’s fine if someone gets a really high grade, like 100%. We have a good idea of what 100% means. I find that basically everyone who gets 100% in my classes is hard working, has a good grasp of the material, and is engaged.

But say someone gets 83% in my class. That’s a pretty good grade. But it actually doesn’t let the student signal usefully. Here is a profile of several students who got 83% in my classes.

Student A was very bright and engaged, but decided to skip one of the exams. She didn’t have any excuse for this, and also didn’t come and retake the exam even though I offered her this possibility.

Student B was really excited about the subject and worked hard but didn’t come into the class prepared, and so had a lot of trouble understanding the material, especially at first

Student C basically understood the material and did well on exams, but was so sloppy with homework that even when he did it, he got most things wrong

Student D was very hard working and able to remember a lot of information, but had trouble with  some very basic concepts.

What do I do with this if I’m someone the student wants to signal to? If I’m an employer? There may be situations where I would, say, strongly prefer to hire student A over student D. Or, there may be situations where I would instead strongly prefer student D. But the point is, they are very different students, and would make very different employees. What information is that 83% giving? My contention is that it gives nothing at all beyond “it’s a pass, and it’s not 100%”

So from the point of view of a student or an employer, these grades are close to useless.

And yet the difference between, say, 83% and 85% drives students to immense amounts of stress and unfortunate behaviour. This week in class, for example, a student gave an angry tirade during class time in which he said that “people’s lives hang in the balance” because he got an 18/20 instead of 20/20 on an evaluation. I’m pretty sure people’s lives don’t actually hang in the balance over two points, and if they did, that would be terrible. But even the fact that students can convince themselves that they do shows how out of whack and stressful the whole arrangement is.

Are grades at least good for me, the instructor?

From my point of view, grades are supposed to be like karma: a system that motivates the students to learn despite themselves. It’s true that the primary motivation for most students in most courses is getting a good grade, so it’s somewhat useful. However, I think the current grading system manages the allocation of that motivation poorly, in that it diverts all motivation into doing make-work projects. Students always want to do things for extra credit rather than do the assigned work. The problem with extra credit is that while it is easy to demonstrate conscientiousness by doing a bunch of extra work, it’s much harder to demonstrate understanding. As a result, students are basically not necessarily motivated to take the time to understand things. Whereas the stress they experience is pretty real

The current grades system also leads to problems like grade inflation and using student grades as teacher evaluation as I’ve talked about recently. Not to mention how frustrating grading itself can be. So it would be ideal if there was some alternative.

I think in a small class setting like I have, written evaluations can make sense. But even something like a three-axis system where I could write “extraordinary / much better than typical student in class / better than typical student in class / typical / worse than typical / much worse than typical / unclear” for conscientiousness, understanding and engagement. Like yeah, I see the problem with this system. It might end up being extremely biased against minorities. So maybe this is also a bad system. But that doesn’t mean the current grading system is a good one.

Posted in INMJTEY, shit we have no idea about | Leave a comment

Teacher Evaluation

One of the biggest, if not the biggest, controversy in education in the US is about standardized testing. And specifically, using standardized testing to evaluate K-12 teachers. Intelligent people, whether supporters or detractors, generally agree that good teaching can be much more than just getting a student to pass a certain standardized test. At the same time, there needs to be a method of evaluating whether teachers are doing a good job.

Where does my teaching experience put me? My students, being post-secondary, don’t generally have to take mandated standardized tests. So I don’t have direct experience with this form of evaluation. But I do have experience with some aspects of “automated” teacher evaluation, which may be generalizable.

I am someone who, temperamentally, can easily fall prey to a kind of wishywashy appeal of stuff like “inspiring” teachers. I’m inclined to romanticize non-quantifiable things, and I know this about myself. I am also a scientist by training. Therefore, I try to be doubly wary of any argument against using quantifiable data.

At the same time, my experience in seeing some evaluation metrics for teachers gives me pause. Of course, you evaluate what’s quantifiable. You evaluate what’s quantifiable because that’s what you can evaluate. But being quantifiable in itself doesn’t guarantee that something is meaningful. And a meaningless number, if it is used to justify important decisions, can be worse than no number at all.

And often, when evaluating teachers, administration uses metrics that don’t seem meaningful to me.

For instance, it is generally agreed that an important principle of good teaching practice is to match your evaluations, what you teach, and what you want the students to learn. This seems to me obviously a good idea.

Part of what is necessary for this matching to happen is to make sure the things you want the students to learn are tangible skills and actions. Because it’s only tangible skills and actions that can be well evaluated. You can’t test the students on whether they “understand Gauss’ Law” for example, but you can test whether they can “apply Gauss’ Law to determine the electric field inside an insulating sphere”. So it’s important that the learning outcomes (what you want students to get out of taking a course) are things you can test.

My school wants to make sure that this is the case for our classes. The idea behind it is sensible. But the method for ensuring this is not, in my opinion. They look to the syllabus for what the listed learning outcomes are for the course. And they count what proportion of learning outcomes have “action” verbs. And if it’s not 100%, they write you an e-mail, telling you to change the syllabus so that it’s 100%.

(And then another e-mail. And then there are several meetings. etc…)

But judging how well your teaching matches your evaluation by the proportion of action verbs in your syllabus is worse than useless. It’s actively bad. I don’t know how much of standardized teacher evaluation is doing something of this kind. But I worry that at least some is.

At the same time, what is clear to me is that teacher evaluation is necessary. My suggestion is that it be done by peers sitting in and giving feedback on each others’ classes. This is of course imperfect, imprecise, and prone to gaming of the system. But I think it’s the least bad method we’ve got.

Posted in INMJTEY, shit we have no idea about | 1 Comment

Grade Inflation

Before I began to teach, I thought most concerns about grade inflation were silly. I still think that, but now I also think grade inflation is a big deal. Because after starting to teach at a technical college, I realized that there’s actually a real systemic problem, and the consequences are worse at lower-prestige institutions like mine. But that’s not what most people talk about when they talk about grade inflation.

Instead, talking about grade inflation comes off as old people grumbling about how easy the young ones have it. I don’t want to hear “kids these days”-style moralizing. As far as I am concerned, if I were in the mood for it, I would just read “The Meditation of the Old Fisherman” as many times as necessary. And I really can’t bring myself to care that an “A” from Harvard now means what a “B” from Harvard used to mean. So why is it that I’m worried about grade inflation then?

I’m worried because at institutions like mine, grade inflation is part of a cycle that wastes students’ time and money, and makes starting their careers more of a hassle. We’re mostly preparing people for certification in trades and in health-related work. And we’re locked in a terrible positive feedback loop with the employers who hire our students, where grade inflation plays a big role. It works like this:

Putting in a bunch of extraneous requirements to be certified for a job is advantageous to the licensing body (this makes the job sound more prestigious and important). It is advantageous to people who are already working in that job (reduces the amount of people who can start working in the field, thus reducing competition). It is also advantageous to the college (more people need to take more classes and pay more tuition). It’s also very hard to push against. Arguing against increased certification requirements makes you sound anti-education (since it wouldn’t be worse overall if, say, radiation technologists were better at physics. It’s just not really worth it for them to go to school for an extra year to achieve that, for example). Thus, the requirements for licensing are continually made more strict.

As a result, a bunch of working class students looking to start careers end up taking courses in things they don’t really need to learn, but which are required for certification.

It’s in my interest, since I am an adjunct and am evaluated on it, to give them the grades they need to get into whatever programs they are going into. But honestly, totally apart from that, I don’t want to give low grades for required courses if I know that what the students are missing is not really required for people to do the work well. It seems unreasonable to stop someone from being an LPN over some arcane piece of statistics curriculum, say, when whether they are able to know the difference between a z-test and a t-test doesn’t really impact how good of a nurse they’ll be.

Or does it? I’m not a nurse, and have never held any job in the health professions. So I’m not actually qualified to answer that question. Luckily, our college’s nursing faculty have a lot of nursing experience. I went and talked to an instructor in our nursing department to find out.

They said that some math is actually really useful for nurses: they need to be able to understand proportions and do unit conversions quickly and reliably. Without that skill, they can’t do their jobs properly. But when deciding which math course would be required by the state for entry into nursing programs, it was decided (unwisely in the instructor’s opinion) to make the class statistics. Because it would be more prestigious, and also be connected with nursing research.

The instructor also gave me a textbook on nursing research. Here is the textbook’s explanation of what you should be doing when you first come up with a study to do:

These activities include reading, conceptualizing, theorizing, reconceptualizing, and reviewing ideas with colleagues or advisers. During this phase, researchers call on such skills as creativity, deductive reasoning, insight, and a firm grounding in previous research on the topic of interest.

Textbooks, man! But that’s beside the point.

The point is that I do what I can to “sneak in” unit conversion examples into my statistics class. But there are also external requirements for what I have to teach — so, I’m sorry, but these nurses will have to learn about how to perform a t-test. But if they can’t do it very well — well, now I’m pretty confident that that’s not really going to make them be worse nurses. Therefore, I am not going to let it stop them from getting into the nursing program.

And yet, as a teacher, I can’t just make up grades — I have to base grades on the work in the class. I’ve seen this having a negative effect on how courses are taught and how students learn. Because I try not to prevent people from going to their program of their choice if they put in some effort and aren’t missing anything that would be crucial to them in the profession, I have to structure a course in such a way that you can get a very good grade without knowing a very large percentage of the material. There are two ways of achieving this:

  1. Not teach any material the students don’t already know
  2. Not check whether the students have learned the material

Both approaches lead to a class focused on busywork and stop students from learning the material in the course. Which I guess would be OK according to my premises, since they don’t need it, but it’s a dishonest system and wastes the students’ time and money.

And not only is it bad, it keeps getting worse.

Employers and licensing bodies see that the grades no longer mean what they used to. Meanwhile, they still want to maintain the prestige of their profession and stop competition from new workers. Therefore, they try to tighten the licensing and employment requirements further, meaning that students now need to take even more classes, and get even better grades in them. As a result, we might be required to tailor our classes even more towards busywork and ensure that everyone can get a good grade no matter their level of understanding. But the range of grades we can work with has shrunk — instead of going between 50 and 100%, we now get to range student understanding between 75 and 100%, and then between 90 and 100% and so on. The students have to do more and more work that requires less and less understanding. They waste more and more time and end up learning less and less.

Since many students have barriers to success — many of them, for example, are single parents who also have to work nearly full time — some of them don’t end up finishing the courses the first time around and can’t enter the profession. Which is great for the licensing body, but terrible for people’s lives.

Of course, this has mostly been the standard story that the more libertarian-minded people of good faith tell about licensing requirements. The point that I want to highlight is that we can maybe combat some of this trend by combating grade inflation.

Posted in INMJTEY | 1 Comment

The Flipped Classroom

This is a thought I’ve just recently started having, as I’ve started teaching in a “flipped” classroom, and I’m not sure if it’s at all true. I wanted to share it in case someone who is more knowledgeable about the world of education and education research would comment. To briefly summarize, a “flipped classroom” idea is that the concepts are delivered to you outside the classroom (e.g., through videos and notes) and then the class time is spent on problem solving and group work, applying the concepts already learned.

My initial thought was that this is a good system for teaching concepts to the median student, but was not effective for the very best and very worst students. I thought that, since this method implies you have to put in work outside of class to be able to do anything in-class, the students who were not putting in enough work outside of class would quickly fall behind in a way that left them unable to catch up, since attending class would be useless for them. However, those who did put in the work would benefit.

Very preliminarily, I’m finding that something like the exact opposite is true. The flipped classroom is not useful for people who are relatively diligent, but really great for people who do no work whatsoever. Since my job as an instructor becomes more tutor-like, I am able to go through problems with people who don’t get them at all. Since the people who don’t do the assigned work will not get the in-class problems at all, I spend almost all of my time with them. As long as they admit they don’t understand what’s happening and are receptive to help, which is obviously not the case for everyone.

As a result, in my implementation at least, the class of student who benefits greatly is “the student who wants to learn but doesn’t put in even the minimum amount of effort to do so”. Whereas the median student basically has very little reason to come to class. This looks good on evaluation metrics (a much larger percentage of the class gets C’s or above than before, say), but I’m pretty sure it’s non-ideal.

Posted in INMJTEY, teaching | Leave a comment

Cost Disease in Education

At Slate Star Codex, Scott Alexander writes a post on a pretty important question: why do things cost so much more these days, without there being any particular benefit to the average person? Like he mentions, there are many people who think they know the obvious answer, but they disagree on what the obvious answer is. For my part, I think the answer cannot ignore the fact that the cost disease is happening at a time of ballooning inequality, something that Alexander mostly elided in the original post. All the same, in my field of education, I think a part of the explanation lies in a combination of Alexander’s points #4 (increased regulation), #2 (things like administrative bloat), #6 (reduced risk tolerance) and #5 (fear of litigation). I wouldn’t necessarily put the negative gloss on this that that implies, though.

Here is how I would tell this part of the story. Our society used to be run in a way that was basically majoritarian. I mean that in the sense that the institutions of society were created explicitly with the majority in mind. With time, though, as a society, we seem to be gradually, fitfully moving towards the idea that institutions of society need to cater to everyone in the society. This is the aspiration that guides naming US education law “No Child Left Behind”, etc. However, this gradual change has come with a huge increase in cost, because accommodation of everyone is disproportionately expensive.

In my day-to-day work as an instructor, I see very clear examples of this dynamic.

For instance, I used to use the whiteboard in my classroom to explain material I was covering. My students would take notes based on what I wrote. However, not everyone in my class is necessarily able to take notes (I’ve had students who were veterans with hand tremors, for example). So, I switched to taking notes on paper, projecting the notes on a document camera, and then scanning the paper afterwards and posting it online so that people in my class can have a record of what happened in class, even if they can’t take notes.

This solved the problem for those unable to take notes, but it created a new one. Any piece of material posted online needs to have alt text so that online material is accessible to the vision impaired. If I made handwritten notes in class, I would have needed to recreate the notes after class in a typed version as an alt-text. Which is difficult to do, because the notes involve a lot of diagrams and mathematical symbols.

Therefore, I needed to go to a third system, preferably one without a lot of in-class handwritten notes. So now I prepare my notes electronically beforehand, show them in sequence, and if something does come up that requires new notes, I have to then type it up after class.

The amount of work involved for me is significantly greater than the original “writing on the board” method. Whereas the benefit to most students is not there. Most people can take notes, and don’t need this procedure. The fact that my notes are pre-arranged has some positives, but also some negatives in terms of flexibility in responding to questions, etc. In addition, since students know the notes are available, they are less likely to take notes. Whether taking notes helps one learn or not is very individual — but I would say on the whole it does. So, the entire endeavour I would guess makes my students less successful on average. It probably is enough to cancel out the on average one student per quarter who needs note-taking help.

Meanwhile, it’s true that people who need one kind of accommodation are not the same people who need another kind of accommodation. In fact, the number of students I’ve had who both requested notes and needed alt-text for online materials is zero.

So, let’s sum up the previous anecdote. The overall results are that I have to do a bunch of extra work, at no net benefit to students. In addition, we need new equipment (a document camera), and tech support (to set up this camera and the associated recording software). We also need administrators from accessibility services: someone needs to interview the students with special needs to figure out what those needs are and how they can be accommodated. Someone needs to explain this to me. And someone needs to monitor that I am in fact doing this.

Scott Alexander can look at the situation and say “well, the overall performance in zolltan’s class seems to have gone down, and the cost has gone up tremendously. The mystery of cost disease strikes again!” — and that’d be correct.

However, if we are to take seriously our mission to educate the public, then it should also be the case that we educate the public, and not just those who we find it convenient to educate. So this is not a case of an obviously bad decision as much as choosing between real tradeoffs.

Is this the right choice? Is it worth it? I tend to think that in the case of public education, it is. But the question is not at all an easy one. And answering it needs the participation of people whom it directly affects most of all — i.e., those students who need accommodation.

I don’t claim that this explains all of cost disease everywhere, or even much of it, necessarily. But what it does for me is put in context the complaints about bureaucratic inefficiency and how private companies can be streamlined while the government just wastes your money. You don’t have to ask whether private companies are more efficient than public ones to figure out why the price of public education is increasing. You just note that, say, educating 95% of the people is a lot, lot cheaper than educating all of them.

Posted in INMJTEY, politics, shit we have no idea about | Leave a comment