Reflections on registration (#1188)

Topics/tags: Registration, Grinnell, long, rambly, footnotable

I’ve been thinking a lot about Grinnell’s registration these days. I suppose many people at Grinnell have. After all, we’ve just reached the end of what may be the last traditional Grinnell registration, and we’re in the cleanup phase after cuts, closes, and balances.

In case you’ve forgotten, or in case you never knew, Grinnell has a different registration system than most institutions: Instead of employing a priority system or a first-come, first-served system, at Grinnell, we permit students to preregister for their classes without class caps; then we shuffle things around to deal with the courses that are overenrolled. Sometimes we cut students from overenrolled classes. We might balance students between sections. We close other sections to hold slots for the cut students. Sometimes we cancel low-enrolled classes altogether and add new sections of high-demand classes. And, once in a while, the Dean coughs up some cash, and we try to find people to teach new sections of high-demand classes [1].

It’s a weird system. However, it has some significant positives. From my perspective, one of the central positives is that it’s somewhat agile; if we see demand for a class, we can shift resources or at least plan resources for a subsequent semester. But it also reflects a central institutional value: We don’t (or shouldn’t) treat our students as entries in a database. For example, when making choices about cutting students from a class, we can often rely on our knowledge about individual students. Will a student be abroad in another semester, making it necessary (or at least better) that they take the course this semester? Is a student completing a complicated double major that creates conflicts? Does a student have an upcoming internship that relies on the course topic? Has the student been cut from the course in the past?

As a member of a high-demand department, I also know that we rely on the system to provide a bit of a balancing act. We don’t have enough slots in CS courses to handle all of the demand; I’m pretty sure that few small liberal arts colleges do. So we try to consider a variety of factors in reshuffling our classes. For example, we prioritize everyone making progress in the major over any individual student getting multiple courses. We also tend to prioritize potential majors over students taking our classes for general interest [2]. For example, a rising second-year undeclared student is more likely to get into CSC-207 than a rising third-year student with a declared major in another subject.

Dealing with all of those conflicting issues creates a non-trivial workload for some faculty and departments. I don’t like the workload or the work, but I’ve always thought it was worth it for the accompanying benefits.

In any case, that’s our system. It’s imperfect, but it’s individualized.

Or that was our system. Things are changing.

Some folks, particularly those new to the system, sometimes find it problematic. President Harris considers it inequitable, as does our new Registrar. What are the inequities? One obvious one is that it provides a kind of favoritism; faculty can choose who to keep and who to drop or who to allow to over-enroll in a class. Unconscious bias can play a role in that. It’s also the case that the students with more social capital (or those who have advisors with more social capital) are more likely to take advantage of the system. For example, they might know to contact individual faculty to receive higher priority. The current system also requires access to hidden knowledge. What courses tend to fill? Who gets priority when courses fill?

To President Harris, these inequities sufficiently trumped the strengths of our system that she charged the Registrar to come up with a new registration system [3]. The first draft was what I’d call a seniority plus fcfs system. That is, seniors first registered for up to nine credits, then third-years, then second-years, then first-years [4]. Each of these registrations was first-come, first-served. Who can click the Register button first for a high-demand course? It’s like a video game. After that, there was a second seniority round for the remaining credits.

Such a system creates a host of problems. For example, rising seniors could claim all of the slots in a course that second-year students need to make progress in the major. In CS, the discipline I know best, those could be slots in a fall CSC-161 (for those who took CSC-151 in the spring) or a fall CSC-207 (for those who took CSC-161 in the spring). Or nonmajors could claim the slots in The Craft of Fiction even though the department might prefer to prioritize majors.

Fortunately, faculty raised enough objections about the seniority plus fcfs system that we have a new draft system. In this system, we do a short round using our historical preregistration system for the first nine credits, add a short seniority-based cleanup round, and then move to seniority plus fcfs for the remaining credits in the third round. That is, we begin with a short round in which we pretend there are no caps on classes and students can register for up to nine credits. The Registrar’s office then quickly applies cut and balance rules to what we hope are a relatively small number of classes. Students who have been cut (or who forgot to register in the first round) have a short cleanup round where they can add classes up to a total of nine credits. Perhaps others can shift things around in that round. That round also uses seniority and fcfs. Finally, we have a third round of seniority-based fcfs registration for the remaining credits.

I think this newer system addresses my primary concern. For example, if seniors register for CSC-207, they will have lower priority than rising second-year students, so the latter will get their slots, provided they know to register for them in the first round. I do worry that some students, particularly double majors, may have more complicated schedules that can be easily broken. And I worry about the few critical course pairings, such as BIO-251 and CHM-221 [5], that need careful arrangement to fit into many schedules. I guess we’ll see what happens for those students.

Is this new system better or at least more equitable? I’m not certain. Students still need some knowledge of what courses are likely to fill. Without such knowledge, how will the know what to select in the first round? I suppose they can see courses filling and shift things around [6], but that assumes they have time to keep track. Students will still benefit from inside knowledge about processes and policies, such as which majors you should declare to ensure priority in certain courses [7,8]. Some students will still ask about waiving prerequisites. Is there inequity in waiving prerequisites? As long as we’re consistent in how we waive prerequisites, there shouldn’t be. However, some groups of students may be more likely to ask faculty to waive prereqs.

I’m also unsure whether a first-come, first-served system is truly equitable. The current draft has 7 a.m. as an opening time for registration. But some student athletes have practices at 7 a.m. and other students work at 7 a.m., most often in the dining hall. In addition, not everyone finds it easy to wake up that early. I don’t know that there’s a better time; as far as I can tell, any time you start registration will cause problems for some students.

On a separate note, I worry about the effects of this system on diversity in my discipline. In national discussions among CS faculty, I’ve heard that experience and studies suggest that when resources are tight, underserved students tend to persist less. Many factors come into play. For example, those with privilege tend to be more willing to push for a seat or to wait things out while those who don’t see many people like them may be more likely to say I don’t really belong and look elsewhere. While we are not permitted to prioritize students from underserved groups [9,10,11], it feels like we need to do more for them than this system permits.

I’m sad about the loss of instructor autonomy in deciding who to take next from the wait list. It feels to me that taking students off the wait list is one of the important times that individual attention comes into play. Perhaps we can be more equitable in asking everyone for their story. But a purely policy-based system removes individual considerations. Of course, I should also accept that instructor autonomy can also lead to inequitable choices, such as my temptation to prioritize students from underserved groups. I also know that my wait lists are usually first-come, first-served. It’s complicated.

I don’t know what the effect will be on horse trading. What do I mean by that? Currently, we sometimes offer incentives for students to give up their spot in an over-enrolled class. For example, If you switch from CSC-341 (overenrolled) to CSC-301 (slots available) this semester, we will not only prioritize you for CSC-341 in the spring but also move you up on the wait list for CSC-213. [12] Is horse trading the best strategy? I’m not sure. However, it is certainly one way that we address problems in making sure that all of our students get courses. It’s also a way of reducing problems down the road; if we don’t fill all of our slots in CSC-301 in one semester, we’ll likely be over-enrolled in the following semester. With approximately sixty CS majors in each class year and three sections of CSC-301 each year, each capped at twenty, we need to fill every seat every semester to prevent future demand that we cannot meet [14,15].

I worry about the difficulty of writing priority lists for students coming off waiting lists. Many factors come into play: Major, class year, study abroad, whether they’ve been on waiting lists in the past, etc. For a fall CSC-207, I’d give the highest priority to a rising third-year declared CS major who plans to study abroad in the spring and who has been cut from CSC-207 in the past. Of course, I don’t know if any such people exist. After that, it would be rising third-year declared CS majors who are studying abroad in the spring. Then rising third-year declared CS majors. (Oh, that’s right; I should separate those previously cut.) Then rising second-year students who have been cut in the past [16]. Then other rising second-year students. Or should that only be those who do not already have a declared major? And then? Perhaps Statistics concentrators who need it for their concentration. Possibly students who need it for their graduate program [17]. What about rising third-year intended CS majors? How do we identify such students? In any case, we will encounter complications and spend a nontrivial amount of time to make and agree upon the priority rules. And I’m fairly confident that we’ll leave off some category. For example, what do we do about the student who forgot to register but needs a course? Where on the waiting list do they go? As importantly, who makes the decision? [18]

There are also priorities I don’t know how we’ll express or record. How do we document (and implement) the policy that it’s more important that a rising third-year student get a course that they need to move forward in the major than that a senior get the course if they don’t need it to graduate? [19]

Once departments have made policies, Grinnell will have to communicate prioritization and other information to students. We’ve not been good about doing so in the past. Can we do better? Certainly. Can we do better by this fall’s spring registration? I’m not as sure.

Of course, I also see positives to the system. Once we’ve done the up-front labor of creating priority lists, we won’t have to spend the effort managing waiting lists. And when students can’t take classes they want or need, we can blame the system rather than try to find solutions. Oh. Wait. That means we’re not giving students the individual attention they deserve.

Having watched the Registrar’s office attempt to do CCBs this semester, I do worry that we’re moving toward treating students as boxes, rather than as individuals. That may be more equitable, at least in some sense of equity, but I’m not sure it’s more fair [20].

Still, I appreciate the goal of equity. And if the new system requires less work, that’s a good bonus, too. In any case, I look forward to seeing what happens. I hope it goes well.

Postscript: I’m glad I’ll be on leave and so will only experience most of the complications indirectly [21].

[1] I don’t think it’s a good idea to rely on contingent faculty. If we have predictable needs, we should have more tenure-line faculty. Sometimes, the best solution is to ask folks in a shared contract to expand their teaching; the Faculty Handbook specifies appropriate compensation (1/5 a year’s salary, if I remember correctly).

[2] I wish that every student who wanted to take a CS class—or even multiple CS courses—could do so. However, the institution appears reluctant to provide us with the resources to ensure those opportunities and, well, no one listens to my complaints about those issues anymore.

[3] We will, for the time being, set aside the governance issues involved in that charge. From my perspective, the President should have charged the Dean to investigate the issue.

[4] I’ve forgotten what the additional rules were for mid-year graduates.

[5] I think.

[6] At least, I assume they’ll be able to shift courses during each round.

[7] I know of at least one rising third-year intended CS major who didn’t realize that they needed to declare a CS major in order to get CSC-207 in the fall.

[8] There are groups of students who are very good at sharing this insider information.

[9] At least I don’t think we are.

[10] I have switched to using underserved rather than underrepresented to emphasize that it is our responsibility to serve these students, rather than their responsibility to represent, as it were.

[11] However, I find it problematic that underserved looks far too close to undeserved.

[12] That’s not exact, but it gives the flavor.

[14] It’s a bit more complicated than that. Some class years have more than 60 majors. Some class years have slightly under 60 majors. Some students can take a CSC-301 equivalent while abroad, although it’s not one of the more common study-abroad courses. Some students need to repeat CSC-301. And, amazingly, some students who are not CS majors want to take 301. In any case, the simplest model is that in CSC-301 and CSC-324 and CSC-341, we need to fill every seat every semester or we will find that we have to cut students out of required courses.

[15] When I started writing this musing on April 10, we still had slots available in CSC-301. When I finished musing on April 11, all the slots in CSC-301 were full.

[16] Those students generally won’t be studying abroad in the spring; most Grinnellians who study abroad do so in their third year.

[17] Yes, some graduate programs (particularly in STEM disciplines) require that you’ve had an undergraduate class in Java.

[18] It’s not just a hypothetical question. I have an advisee who forgot to include one course on their schedule. They are now on the waiting list for that course. I hope they are prioritized.

[19] And do we even agree about that policy?

[20] Now that I think about things a bit more, I find that the system may be equal but I’m even less sure it’s equitable. I thought equitable systems meet people where they are, which requires taking individual situations into account. Isn’t that what the famous baseball soapbox cartoon suggests?

[21] Perhaps I should have given up my advisees [22]. Then I’d get to watch from even further afar.

[22] Perhaps I should not have taken on more.

Version 1.0 of 2022-05-11.