Two summers of student-faculty research

This was my second summer of participating in Whitman’s student-faculty summer research program. Below I share my stories of two projects that were quite different in their content and their material circumstances.

I was on the fence about taking a research student last summer. Emma Twersky ’17 read my December 2015 blog post and volunteered herself. Never before had I conducted summer research with a single student; I had always supervised a team of at least two. A well-functioning team of students can support each other and hold each other accountable, and can actually require less supervision than a single student. But I was willing to give Emma a try as my only summer research student, both because I knew her fairly well and because the analytical project I had planned was not work I could delegate but work I would want to be closely involved with.

It’s just as well, as Olin was under construction and it would have been hard to find a place to work near a team of students. Fortunately, Emma was very independent and able to get things done while meeting only a few times a week. We often met over lunch or coffee, and even once at my home, to work through our analyses together.

I thoroughly enjoyed our conversations. I do not think I had ever collaborated with a student on that type of work before; in fact, I hadn’t done any work like that since I was a student myself. It was so much more fun than doing it by myself would have been. And our meetings didn’t just hold Emma accountable for doing the work; they held me accountable as well. I’m not sure I would have got it done on my own.

At the start of the summer, we considered three possible publication venues. We narrowed it down to PERSUASIVE 2017 largely on the basis of the calendar: the other two conferences fell over Emma’s graduation. Nonetheless, it was a good choice. It’s a conference I had already attended several times and presented several papers, including one with students. I was familiar with the audience and I knew what to expect from the review process. And while this was not the most high-profile of the conferences we considered, it was the audience I most wanted to reach.

Our goal was to have a complete draft of the paper by the end of the summer. Emma held up her end of the bargain. I gave her copious feedback on her sections, even apologizing for what felt like excessively detailed criticism on style and usage, because I wanted to be sure it would be up to my standards for publication. We had a complete draft of her sections of the paper by the end of our 8-week timeline.

In the fall, Emma transferred the paper to ShareLaTeX (which I had not used before) and spent a significant amount of time getting the tables and figures into shape for submission, while I wrote my sections. It took several rounds of revisions to get the paper down to the required length, and we were lucky that the paper submission deadline was extended by several weeks. That said, Emma had done good work and I was confident the paper would be accepted.

We jointly presented the paper at PERSUASIVE 2017 in Amsterdam in April. We met a couple of times in March to prepare our presentation. Emma discussed several examples from the paper, since she knew them better than I did, while I provided the overall frame. I’m glad we presented our work at PERSUASIVE—it was indeed the right audience.

If you’d like to learn more about the content of my research with Emma, here’s a link to the conference paper. I’m proud of what we accomplished, and I may further develop the work into a journal article.

This past December, the CS faculty instituted an annual summer research info session. A long-standing institution in my prior department at Grinnell College, the main purpose of this session is to inform students interested in summer research about planned faculty research projects. (I also took the opportunity to put in a quick plug for national research programs, including the NSF REU program and the CRA DREU program, as well as national research labs, such as PNNL which is quite nearby.)

Andy and I pitched a joint project, in which my students would develop a user interface using persuasive system design features, and his students would develop back-end natural language processing algorithms. One student followed up with me: Andrew Harvey ’20, who worked with me this summer. Two students were slated to work with Andy. I was grateful that Andrew would have two other students working with him on the same project, even if they were working on different aspects of the project. Unfortunately, that was not to be; both of Andy’s students, for different reasons, ended up not following through on their proposal for summer research at Whitman.

Although I fear Andrew got a bit lonely at times, it worked out all right. We ended up doing more of a straight user-centered design project than the research project Andy and I had originally envisioned, which in retrospect was a more appropriate place to begin the project and also more appropriate to Andrew’s prior experience as a rising sophomore. Since Andrew needed someone to work with on ideation and evaluation, I got to participate in those activities, a nice refresher since I teach them in my HCI course but hadn’t done them myself in several years. Andrew became an expert on using Thunkable as a mid-fidelity prototyping platform, and was even awarded New User of the Month for his contributions to the Thunkable forum.

I was in and out of the office more than I expected, both because of time spent moving houses  and for other personal reasons that will be the subject of a future post. Fortunately, Andrew kept very regular work hours, far more regular than mine, up to the end of the project where all he had left to do was write the technical report. Since he worked in the CS Commons just outside my office, it was easy for us to coordinate with each other whenever I was in the office. We scheduled a weekly kickoff meeting to set goals and (at the beginning) plan tasks for the week; a mid-week check-in which we often skipped because we were already talking about work in progress; and an end-of-week meeting to review progress and consider directions for next week. Since John also had two summer research students, we had an informal student-faculty lunch on Tuesdays and weekly presentations or progress reports on Fridays, which lent some additional structure and social contact.

We made a last minute decision to start summer research a week earlier than we had initially planned. Our proposal stated we would start a week after Commencement; instead we started the day after Commencement. There were two reasons for this. First, Andrew needed to take a week off midsummer for a family gathering. Second, I thought it would be helpful to start at the same time as John’s students, to start our weekly lunches and presentations right away and maximize the students’ overlap during the summer. I think this worked out much better for Andrew than our original plan would have.

It was also good for me in that we still ended the project at the end of July, even with Andrew’s week off, and another day and a half off to participate in Code for Good. That meant I could tag along on my husband’s August work trip to the UK with a clear conscience. On the other hand, it would have been nice to have some time between Commencement and the start of summer research to wrap up the spring semester and make a clear start to the summer. I never did process my spring course evaluations, put spring semester materials away, or set professional goals for the summer. A musing from mid-summer reflects my dissatisfaction with not really feeling on top of my work.

I framed Andrew’s technical report as providing history and guidance for students who may continue the project in the future. I am also encouraging Andrew to submit a poster proposal to CCSC-NW, which will be quite close by this fall. I think we’ll pick that up soon after classes start.

Unfortunately, the final form of the project does not lend itself easily to a research paper. It might be possible to develop an experience report as a short paper for PERSUASIVE 2018, since there has been relatively little written about user-centered design in that venue. But most likely I will let that go, as I am having a hard time thinking how to frame it so that Andrew could contribute, and it will be difficult for me to travel in the spring in any case. However, we got enough done that I feel confident it would form the basis of a solid senior capstone project, perhaps co-advised between me and Andy. That capstone project team is the likely audience for Andrew’s technical report. So, there is a positive outcome, just not the one we initially proposed.

I’m still on the fence about whether I will take research students next summer or take a summer off. The most important milestone I need to complete next summer is my proposal for a sabbatical in 2019-20, so as I begin to outline my proposal I should also think about how work with students could help me prepare for my sabbatical.








Leave a Reply

Your email address will not be published. Required fields are marked *