Reflecting on the CSC 151 project at my fourth Obermann seminar

Topics/tags: CSC 151, teaching, academia, long

Every two weeks or so, the Obermann Fellows meet to discuss work in progress at a bi-weekly seminar. As I mentioned recently, I decided to use my second presentation as an opportunity to get feedback from a very different group of faculty about a possible final project for the new digital humanities version of CSC 151. Preparing that project description also encouraged me to think about an alternative and to develop a rubric for the project. I’m recording some of my observations in this musing so that I don’t forget them [1].

To provide some context for the material, I had written an introductory letter that ended up being about two pages long. As I prepared for the meeting and had a hallway conversation with one of the Fellows, I realized that I needed a bit more background. So, in introducing the discussion, I talked about the tight restrictions on the old MediaScheme project (your procedure takes exactly three integer arguments, two of which are width and height; it must scale and stretch precisely), the comparatively loose restrictions on the data science project (find an interesting dataset and do something with it), and some of the data sets that the data science students came up with (alcohol sales in Iowa, incarceration data, cancer data with predictions, poems).

Given the way I framed the discussion, much of our initial conversation was on the two models. The consensus seemed to be that the more focused version of the project was more appropriate than the open-ended version. However, some raised the concern that the number of different projects would be limited. I’m not that worried; I’m hopeful that one of the outcomes is that the students realize how many different things that you can do with the same archive. Some may write programs that visualize parts of the archive. Some may write programs that categorize items in the archive. Some may write programs that summarize items in the archive. Some may use the archive as a source for generating new material [2]. And, as I tried to suggest in the writeup, each kind of programs admits many different approaches. One can visualize connections in the abstract or locations on a map. Different groups will likely choose different categories and different ways to categorize. Summarize can be interpreted in many different ways. Or at least that’s my fantasy.

Still, there were some useful suggestions about handling the topic that I need to consider more fully. For example, it was suggested that even if I make work on this archive with this audience the default, but also allow students to choose more broadly, Those who are willing to spend extra effort can choose their own archive but will need to discuss that archive with me first. Another suggestion was that I give them two or three very different options so that they would be more likely to find one of the archives of interest [3]. I like those as long-term suggestions. However, for the first offering, I am inclined to work with a single archive, preferably one that has a real audience. If I’m providing the archive, I need to spend more time getting it into the right shape (or at least understanding it), which takes time. A uniform archive also helps support uniform grading. And, that way, I can bring a real client to the class. Fortunately, my fellow Grinnell Digital Bridges fellow has an archive that they volunteered to let us use. We need to talk more [4].

I got a useful suggestion for the times that I am picking the data set myself, or giving students an open-ended option, and that’s to use ICPSR. The Fellow who suggested that resource did note that it’s way too easy to lose yourself in those data. But it does strike me that ICPSR can provide some exciting opportunities for a humanistic approach to social-sciency data [6,7].

We spent some time discussing the length of the assignment and related issues. If I count right, each student spends ten hours over two weeks on this. How much can they really accomplish in that amount of time? I said that while that is, indeed, a worry, I worry more about student workload. If we assume each four-credit course is twelve hours per week, we are already asking students to spend 48 hours per week on their classwork, and that does not count their work time or their extracurriculars. If I have students in class for four hours per week, give three thirty-minute readings per week, expect them to attend a one-hour mentor session, and give them three ten-minute lab writeups, that leaves five hours per week for the project. I was glad to have my Grinnell colleague note that Wellness is a real concern at Grinnell [8].

I’m not sure where in the conversation it was, but we also ended up talking about preregistration and advising at Grinnell, primarily in terms of the workload. I noted that I needed to find time to meet with my forty advisees [9]. My Grinnell colleague mentioned that many of their advisees ask for hour-long preregistration appointments [10]. That led to a comment on the order of I’m often envious of what you have at Grinnell [11,12], but not after hearing that.

Where was I? Oh, yes. What can students do in ten hours? A whole lot, it seems. More precisely, I think I suggested that,

It depends a lot on the student or group. Some groups will take a previous homework assignment and extend it for this new archive. Some will be more ambitious and start something new from scratch. And, I admit, some will get caught up enough in the project that they will spend significantly more than five hours per week on the project.

That is, most groups can produce relatively substantive projects.

My colleagues also provided a variety of other comments and suggestions that will require me to reflect a bit more on the project. One colleague asked whether it would be possible to have the proposal due earlier so that they could get design feedback, say from another set of students with more design experience, or from students in the client discipline. It may be worth revisiting the schedule a bit. However, if I recall correctly, we don’t cover the last bit of knowledge useful for the project until the class before the project kickoff. In any case, this goes on a list of things to think about.

This suggestion also led to the broader question of whether there can be multiple projects across the semester. I don’t think there’s time for that, but I do think of some of the weekly homework assignments as a kind of mini-project. After I’ve taught the course a few times, I’ll return to the question of multiple projects.

I was also reminded to emphasize the issue of missing information. That is, every collection of information [15] is created by human beings who decide what to gather and what to store about each thing they have gathered. I’ve tried to raise that issue when I teach the data science version of the course [16], but it’s worth raising again in this context.

I asked for language suggestions on the project description. I didn’t get any particularly helpful ones. The consensus appeared to be that, Like all that you write, it’s warm and inviting [17]. I guess I can accept that as my central rhetorical approach.

I also did not get a lot of feedback on the rubric, other than that it seemed quite detailed [18]. Of course, one benefit to presenting your work is that it encourages you to think differently about the work. In the discussion of the project and the rubric, I realized that I might update the project specification to require that students use particular algorithmic techniques (e.g., You must write at least one significant recursive procedure) and include that requirement in the rubric. I don’t recall whether or not we discussed that idea.

We ended up discussing issues of inclusion in CS classes, the benefits of being able to teach a small, workshop-style, introductory CS class, and the likely challenges the CS folks at UIowa folk given their demand and class sizes. Perhaps at some point, I’ll be able to think about how I might scale the FunDHum approach. But that’s a task for another year.

I’m a bit relieved to be done with my presentations. It’s a bit nerve racking [19] to present to the group, especially since I feel like my work is so different than my colleagues’. At the same time, I will miss the need to prepare for a presentation and what I learn from doing so.

[1] I’ll admit that I had more fun participating in a discussion of a colleague’s work; I’ve written about that discussion in a separate musing.

[2] That last approach may not be appropriate for every archive.

[3] E.g., You can choose a collection of translations of Beowulf, this archive of materials relating to mental health advocacy, or Rocky Horror FanFic.

[4] I suppose we could apply for a Faculty-Faculty Tutorial to work out the details. However, I’m not sure that either of us has time for a Faculty-Faculty Tutorial [5].

[5] Conveniently, the deadline for Winter-break faculty-faculty Tutorials has just passed.

[6] Or, more precisely, a computationally supported humanistic approach to social-sciencey data.

[7] If I recall correctly, my colleague noted that the archive includes a record of all lynchings in the US. Such a data set could lead to very powerful projects that might influence opinions and actions.

[8] Our students are very fortunate to have a wide variety of people who advocate for wellness, who encourage students to pursue healthier approaches, and who support them as they adopt those approaches. Still, wellness remains a major concern for students, faculty, and staff. I expect that our new Wellness Director (or whatever we’re calling the position) will also make a positive difference.

[9] That’s not quite accurate. I currently have forty-two advisees. I don’t have to meet with them all. Three are abroad; we will meet via email. Two are going abroad in the spring; we’ve already discussed their classes. One just declared the major; we planned their classes while putting together their four-year plan. Some prefer to communicate via email or casual conversations in the hallway. So perhaps the correct statement was that I had to discuss spring courses with about forty students.

[10] While I have many advisees who ask for thirty minutes, I can’t recall any who started out asking for an hour appointment. There are some who need multiple appointments, but we usually start with fifteen or thirty minutes. I also do enough casual advising and four-year planning that, for many students, preregistration is comparatively easy.

[11] I’m not sure what particular thing that speaker was thinking about. Most likely, our students. One of the big themes I heard at the Digital Bridges was how special the average Grinnell student seems to be. We’ve also talked about the intellectual climate at Grinnell, and the feeling that most faculty feel like they should understand disciplines beyond their own. Or perhaps they were just reflecting on more practical issues, such as class size or our generous sabbatical policy.

[12] I see from my notes that there was a later discussion in which the UIowa faculty noted that they appreciated how carefully the Grinnell faculty thought about teaching. Perhaps that was the source of the envy; we have the time and encouragement to think about teaching and are rewarded for doing so [14].

[14] It appears that the UIowa faculty also care deeply about teaching; there are just different institutional reward structures at an R1 and a SLAC.

[15] A colleague suggested archive. I very much appreciate that suggestion.

[16] I’m not sure that I always succeed.

[17] I believe I’ve already made a snarky statement about the applicability of this comment to my rants.

[18] I was able to achieve that level of specificity by building on the careful work of Janet Davis and Jerod Weinman.

[19] When I wrote nerve racking, my subconscious said You just read a Candorville strip that discussed the use of nerve racking vs. nerve wracking. You should check. But my subconscious was wrong. The strip in question actually indicated that racking my brain is correct and wracking my brain is not. But it appears the same word applies to nerves and brains.

Version 1.0 of 2018-11-07.