I was a panelist at the Doctoral Colloquium (DC) at the 11th International Conference on Ubiquitous Computing (UbiComp 2009) last week, the first time I've participated in that particular track of a conference. The goal of the DC is to give graduate students working on their doctorates (PhDs) an opportunity to present their dissertation work to a panel of external experts - external to their thesis committees - and get feedback from them, as well as the other student who are participating. I want to share some notes from this event, and will write separate blog posts about a UbiComp 2009 workshop I attended and the main UbiComp 2009 conference sometime in the near future.
Each DC student submitted a proposal that was evaluated by two different members of the panel, which was composed of three people from academia (Gregory Abowd, Tanzeem Choudhury and Hani Hagras) and three from industry (Tico Ballagas, A.J. Brush and me). A subset of proposals was accepted for inclusion in the DC program, and those students were assigned a mentor from the panel, and allocated a 30-minute slot during the event, which included a 15-minute presentation and a 15-minute discussion. Thanks to a grant from the National Science Foundation, the DC Chairs (Andreas Butz and Raja Bose) were able to invite all the students who submitted a proposal to attend the event. I suspect - and hope - that all students benefited from the discussions, as many of the issues raised by discussants applied more broadly than to the particular presentations in which they was articulated.
I certainly benefited from the discussions. It was interesting to see the multidimensional diversity represented by the students and their work, with respect to topic areas, approaches, and states of progress in their respective dissertation processes. It was also interesting to see the kinds of feedback the students received, from their designated mentors, other panelists, and other students participating in the event.
Since I view the DC as a forum in which openness - and the attendant vulnerability - is paramount, I will not reveal any specifics about the presentations or specific feedback on any particular presentation. Instead, I will focus on some of the general themes that emerged during the event, especially those that were mentioned during a more general discussion about the dissertation process at the end. In a forum specifically focused on feedback, managing the feedback was itself a challenge. I found myself being uncharacteristically reserved during most of the presentations, taking careful notes and planning to followup individually with a few of the students via email.
One of the most common areas of feedback during the sessions had to do with the generality / specificity spectrum of the proposals. Students who presented rather broad or abstract ideas about their dissertation topics - or a large number of more specific ideas - were encouraged to identify one specific area (or a very small number) within that scope on which to focus for their dissertation. Some of the students who had identified a very specific area - or a few specific areas - were encouraged to provide a broader context for the work to help those outside those areas better understand the potential contribution of the work. In addition to issues regarding the specificity of the work itself, another factor that varied across presentations was the specificity - and measurability - of claims and anticipated contributions. Ideally, a PhD dissertation in UbiComp addresses a large, compelling problem area by identifying an important subproblem, proposing an approach to solve that problem, implementing the solution, providing evidence for the efficacy of the solution, revealing shortcomings and discussing lessons learned and future implications.
During the last session of the day, the panelists were invited to share lessons we learned through our own dissertation processes ... and how our careers have unfolded after the completion of our doctorates. Some of these revelations were rather personal, and so in order to strike a balance between sharing information that may be useful to others and preserving the privacy of the panelists, I will not attribute any of the lessons to any panelist ... except myself (a person whose privacy I rarely protect very diligently on this blog).
My three lessons learned had to do with openness and vulnerability, essentially, being open to
- Change and possibilities: I changed my thesis advisor and area of research midway through my graduate studies, and although it lengthened the time to completion, I think it was the right thing to do; since completing my doctorate, my research focus has continued to change and evolve ... and I have probably had as many changes in institutional affiliation as the rest of the panelists combined.
- Assistance: the members of my thesis committee were well-known - infamous, perhaps - for making significant demands on the students whose dissertations they supervised, leading to less-than-positive reputations among portions of the graduate student population; however, I believed they were genuinely trying to help me create a better dissertation - rather than [simply] creating obstacles - and found that their assistance ultimately proved very valuable. Since completing my doctorate, I continue to be almost pollyannaish in my belief that others who may initially appear to be creating obstacles in my career progression are simply offering assistance in refining or adjusting my path (and this recognition sometimes only emerges retrospectively).
- Letting go: part of the assistance I received was an adamant refusal on the part of one committee member to accept a "claim" that I thought was the most interesting part of my dissertation work, as I could not find a way to effectively measure or assess it (and so it was merely anecdotal); I finally let go of that chapter, and now find that I can't even remember what it was I was holding on to (I think it had to do with the value of definite articles in resolving inter-sentential references). I have since had numerous opportunities to practice letting go of things I previously thought were absolutely essential.
In preparing for a doctoral dissertation I have always found that plans are useless, but planning is indispensable.
I don't want to make too much of the analogy between doing a dissertation and launching the largest amphibious assault in the history of the world (although some graduate students may refer to the day of the Dissertation Defense as D-Day) ... but while I'm on the topic of analogies, I'll mention two other analogies that I referenced in my discussion of the dissertation process. One is to pregnancy: the thesis defense is somewhat like giving birth (from what I've observed), in that it can be painful, but by the time one reaches that stage, one is often so sick of working on it (being pregnant) that the pain of the transformational event seems less burdensome than the growing discomfort of continuing on. [Update: Andy Sack recently wrote about the parallels between a startup and a newborn baby, offering some interesting contrasts between entrepreneuria and academia ... and between starting something and finishing it.] The other analogy is the stages of grief - denial, anger, bargaining, depression, acceptance - that one often progresses through in dealing with a significant life challenge; although this model was originally developed in the context of death and dying, I do think it applies to the Ph.D. process - and other practices that involve dark nights of the soul - as well.
Anyhow, moving on to other, potentially more useful lessons learned and shared by participants in the doctoral colloquium panel, here a few of the nuggets of wisdom shared by my learned colleagues:
- Searching for related work is a continuous, incremental process: If you are working in an area of interest to others, then new work related to your dissertation will always be coming out. And, if your dissertation topic evolves to include new dimensions, then that evolution will open up new areas in which existing work should be explored. The point is not to compartmentalize this part of the dissertation process, or wait too long to begin it, as the exploration of related work may reveal ways you will want to - or have to - [re]define your own work.
- Distributed mentorship: Several people on the panel did not receive the level of support they had hoped for from their primary thesis advisor, and thus sought out advice and guidance from other people - other thesis committee members, other faculty and/or other students. One panelist recommended this process of seeking guidance from numerous sources - possibly changing over time - as a lifelong (or career-long) strategy.
- Collaborative research: The PhD dissertation is intended to represent a single individual's [initial] contribution(s) to the knowledge base of a discipline. However, just because the dissertation itself is a single-author document does not mean that the process of the research has to be entirely a do-it-yourself, solitary endeavor. The research itself can be conducted in the context of a larger effort in which a number of others participate ... reminding me of earlier ruminations on the balance between self-reliance and interdependence.
There were, of course, a number of other helpful insights and experiences shared during the day - too many to comprehensively enumerate here. I'll finish off by thanking the chairs, the other panelists and the students for an enlightening day, referencing a presentation with more nuggets of wisdom (and humor) on The Art of Doing a PhD that was prepared for the UbiComp 2007 Doctoral Colloquium by Jakob Bardram, who chaired the track that year - and is co-chairing the entire conference next year (UbiComp 2010) - and inviting others to share any additional insights and experiences on the dissertation process in comments on this post.