Michael Ekstrand

Affording Quality Conversations

Published on 16 Jan 2017 and tagged with academia.

As professors, we get a lot of e-mails. One of these classes of e-mails is the prospective student e-mail, which comes in wildly varying quality. This takes time from faculty, enough that many have written guides to e-mailing or specific requirements for e-mailing them and receiving a reply; I have a version of this myself.

The standard advice is to be specific about your interest in the particular faculty member you are e-mailing, and demonstrate at least an awareness of what it is that they are working on.

However, there is a flip side to this: if we are going to expect cold e-mailers to be familiar with certain information about us an dour work prior to e-mailing, then we need to make that information clear and accessible. If we don't, it's a bit like the planning commission filing notice of intent to demolish Arthur Dent's house in a disused filing cabinet in an inaccessible basement, though perhaps with less dire consequences.

This is part of a broader problem: the user experience of post-graduate academia is mystifying to newbies. One of the things I have learned reading discussions on Academia Stack Exchange is that a lot of our norms and practices are opaque and unclear. For example, students have difficulty interpreting fairly standard responses to graduate inquiry (see here and here). I've sent responses like these myself, with good reason — I don't want to say much about a candidate's suitability until the grad committee has screened them and I can see the whole application, including reference letters — but the result may be a baffling experience for the recipient.

So, we need to think about what we can do to improve this experience all around. I believe this is a matter of fairness, not just convenience: the more barriers we maintain between students and education that are only breachable by those with the relevant obscure knowledge, the harder it is for prospective students without a particularly strong academic cultural background to break in.

That starts with making the information accessible and easy to find. This is an ongoing, iterative process. Here are some of the things that I am trying to do:

One other thing I am trying is avoiding keyword lists in my research interest statements. I do not know if this will be helpful or not, but whenever possible, I prefer to state my research interests in descriptive prose rather than a bullet list of topics. There are two reasons for this:

  • The topics I study (HCI and artificial intelligence in particular) are extremely broad, so knowing that I work on them is not very helpful for knowing the particular kinds of things I am working on.
  • There is a species of semi-personalized e-mail whose introductory paragraph reads like a copy-and-paste of research interest keywords, and this approach to research interests does not afford those e-mails as easily.

There is a balance here, because there is value in being discoverable by keyword. If a student is interested in artificial intelligence, they need to be able to find the faculty working on artificial intelligence. I think that our Ph.D program's faculty list is a promising approach to this problem: keyword-based categorization of faculty, leading to individual profile pages where we can be more specific.

I do not yet have a good sense of whether or not this is effective, or if it has other undesirable effects. I welcome opinions or data on that issue!

But I do believe that we need to spend some time thinking about the first contact user experience, particularly for prospective students, and work to make it easy to have productive conversations.