Professoring: My Job Is Spreadsheets

Talking about the other parts of campus life

A while ago, a Substack talent-development rep very kindly spent some time giving me some great advice about how to do Substack right. This included:

• posting regularly

• leveraging my research to give people insider views of my work

• posting regularly

• developing some hook for paid subscriptions

• posting regularly

and I took notes and then utterly failed to do any of that. Welp. Sorry! But I took good notes.

Anyway, I’ve noticed that there’s a few types of posts that you folks tend to like, and they tend to be ones about what goes into being a modern professor. So let’s take a look at what I’m doing this week in my real job—and why it explains how so little has changed on campus even after the advent of the Internet, computers, the cloud, and all that techy-tacky stuff.

---

I’ve mentioned before the meme that all white-collar jobs are either emails or spreadsheets. Professoring involves a lot of both. Today, my job was a lot of spreadsheets.

Years (oh, no, decades) ago, I remember that my teachers and many of my professors kept their grades in a literal gradebook. The confident ones used pen. That was never my style—from the moment I started leading my own courses, the workflow was all digital. And that meant…spreadsheets. With sometimes really complicated formulas.

How complicated? Well, it’s easy to grade a course with, say, two midterms and a final, right? Sure! Just decide the weights and write up a formula. Bing, bang, done.

Okay, here’s the thing. Nobody¹ designs courses like this anymore. So instead gradebooks look like this:

In one of my courses, I have

• weekly reading assignments (the best way to encourage folks to do the reading!)

• participation grades

• up to three essay grades

• two midterm grades

• final paper proposal

• final paper

and probably something else, I don’t know. That’s, what, more than 20 columns? That’s one hell of a formula. And when you have one hell of a formula, you know what happens? You make errors. Hundreds of millions of dollars have been lost because of Excel errors. I’m not going to make a hundred-million dollar formula error. But I could screw up grade calculations, and that’s the worst feeling ever.

The easiest way to guard against error is fixing your process. So over time I have put more and more of my process into the Learning Management Systems that my employers use. LMSes are integrated solutions to provide course readings, documents, videos, sound, assignments, quizzes, gradebooks—everything. By keeping everything in one place, there’s less chances of fat-finger errors, lost hard drives, corrupted data—you name it.

There’s only one problem. Mostly LMSes suck. Really. They are all awkward fits. I know for a fact that many of the firms that make them (and these are multibillion-dollar companies) hire a lot of people with classroom experience, but the LMSes all still end up with the overall aesthetics and user experience of the Homermobile. Many of them are quite ugly. None of them offer the flexibility and ease of Excel formulae when it comes to calculating grades. For instance, here is a real formula I had to create to implement a straightforward course policy.

A real formula that really got tested and really works.

I had a series of low-effort, low-stakes assignments. Some were graded on a check-plus/check/check-minus/fail basis, but which would count for full credit if they got anything at check-minus or better. Others were quizzes that would count for full credit if they got 4 or 5 out of 5, but zero if the student got a 3 or lower. (People could retake these.) Finally, I would drop the lowest couple of scores.

This is a very easy course policy to explain. It offers a lot of feedback to learners really fast. And it is an absolute bear to set up in Moodle, which we used at the time. Once it’s set up, everything goes faster than an Excel solution or a pen-and-paper (God) solution, and you can even one-click to send grades to the registrar (so long as you have the conversion from course scores to grades set up correctly!), but it is not fun to set it up.

Canvas, which we use now, is somewhat easier in many respects—but not in all of them. And even when you’re doing easy things, with more than a hundred students, things start to get complicated again. So I found myself today doing some pretty simple work that involved lots of iterating between downloaded spreadsheets and imported grades to get part of the gradebook ready (a full week before grades are due!).

These are the reasons why professors’ eyes roll back so hard when people talk about how easy it is to be a professor and how nice it is to think great thoughts all the time. For easily forty percent of my workweek, my job is indistinguishable from any white-collar job: symbolic manipulation, process and compliance, HR stuff, office politicking, you name it. For the other sixty percent, I’m struggling to get students to learn (somewhat successfully, this term) and to do the work that does the most to advance my career—the research which my line of academia values (a lot, in monetary terms, even if any individual work product is unpaid). And when any of the normal white-collar stuff blows up, like a spreadsheet error, it immediately undermines all the rest of it.

Canvas shell for my US Foreign Policy course this term, entirely hand-made.

Indeed, there’s somewhat less support for me running these kinds of routines than there are for comparable professions. I have to design my courses from scratch—not just the syllabi, but every byte and bullet point that gets served to students (not a few of whom assume that I’m just taking it from something provided by the textbook publisher). I do all my own spreadsheets because we only have support for advice and technical issues, not for taking something scribbled on a napkin and turning it into a course. And although I have TAs for grading, they have, over the years, turned out to have even less understanding of LMSes than I do—neither they nor I receive any official university training on how to use these tools, even though we are in fact expected to use them. (This isn’t something that is limited to my current employer—all of them have been like this.)

If it seems like there’s a lot of reinventing the wheel—well, there is. The same problem has been solved by hundreds of faculty working independently—and never sharing their solution. And if it seems like there’s a lot of friction and wastage—well, there is. The independent sage (and sometimes apprentice) model worked really well in the analogue era, but it fares poorly in the conditions of modern informatized education. On the one hand, there’s CS profs (okay maybe some physics profs too) who skip all this stuff and code everything by hand; on the other hand, well, I have a few colleagues who consider email to be a grotesque intrusion of technology into their pristine lives of the mind, much less learning how to use SpeedGrader.

It’s telling that all of what I’m describing is still directly analogous to, even practically skeuomorphic of, the paper gradebook. Where’s any of the dynamism or customization that computers should enable? Partly, the resemblance comes from the task being the same—turn these humans’ diverse performances into a standardized set of marks—but is there really nothing we can do differently?

But note that the real problem has nothing to do with the technology and everything to do with the insufficient resources geared to enabling the technology to be used well. The inefficiencies multiply, like a flywheel of subcompetence. There’s insufficient support for training professors, so we all use our little kludges; the software is poorly optimized, so we all adjust what we do to what the tools will let us do; and fundamentally there’s not enough time (or incentives) to break the cycle.

You can imagine an organization that reshaped itself to view these sorts of inefficiencies as logjams to be broken—but it would be one that made a big push in the sorts of ancillary efforts that appear on other people’s spreadsheets as costs, and would have almost no detectable (and I use the word very precisely) effect on revenues or retention (of either students or faculty). And so we muddle through, leaky boats against the current.

1. Don’t start.