visible artifacts

By: Chris Malek

Nov 25 2008

Category: Articles

No Comments »

One thing that I acknowledge about myself that I have to break things down into manageable chunks in order for me to be able deal with them (although I have to rediscover this every time I do something new).  I mean that emotionally (a large project can be paralyzing when I think of it as a single coherent unit) and conceptually (especially in the beginning, I find it hard to keep all in my head the amount of information I need to write a paper).   This blog is one way that I acknowledge this about myself: a long blog article (1000-1200 words) is a good managable size for me, and I actually make more progress in setting these tiny projects for myself (”write a post about why I’m interested in visualization,” “write a post about contributor motivation in OSS”) and completing them than I would if I operated solely at the paper level.  My hope is that what I write here in the blog will be useful to me when I come to write papers, because I can build the paper with chunks from the blog.

One of my top fears in my Ph.D. program (after the screening exam) is the dissertation.  I’ve seen people take years to complete it, or not complete it at all and thus never graduate.  I think constantly of how I am going to make the work of it more manageable so that I don’t get to the point where I’ve completed all my course work and then I face this GIANT PROJECT, and potentially become ABD: all but dissertation.   Various faculty have told me that you want to do the smallest project that still qualifies as a dissertation; nobody is expecting you to do otherwise.  But even with that advice, the dissertation is still a terrifyingly large project.   I need a way to break it down into manageable chunks so that I can complete it.

The modular dissertation

What I have always wanted to do is publish a series of smaller, more easily accomplished and digestible papers, and then tie them together with some glue text and call it a dissertation.   I saw several people use this model when I worked for Mark Konsihi in his lab at Caltech, and it always impressed me with how efficient it was: you get published early and often during your studies, and you get to avoid the long slog of writing your dissertation from scratch at the end of your studies.  I always like one action to serve multiple end goals in my working and personal life, and I appreciated how those students did the same in their academic lives.   The other great benefit you get from doing it this way is that you get real feedback along the way, as well as experience in doing science and getting published.

I’ve seen this model of building your dissertation out of papers work in several different ways:

• The first way is what I’m going to call the Yellow Brick Road method.   Just as Dorothy in the Wizard of Oz knew where she had to go (the Emerald City) and how to get there (follow the yellow brick road), in this method you know that answering your research question can be answered by the union of the results from several distinct and separate sub-questions.  First you publish your lit review, then you do a study to address sub-question A and publish that, then do one to address sub-question B and publish that, and finally glue those three together into your dissertation.   For this to work for you, you must know where you’re going and how to get there from the start.  This is very appropriate for explanatory research, because you’ve got to have some serious theoretical foundations in order to do it.
• The second way is what I’ll call the 1001 Nights method.    You do a number of small studies which don’t necessarily build on each other but which are thematically related, and then join them together into a dissertation.   The glue text you write would set your individual studies in a larger context.  Fernanda Viegas (an information visualization luminary) used this method quite effectively in her dissertation at MIT (Viegas 2005).   I would think that this method would work well in exploratory research: we don’t really know how to explore this area, and we don’t know what the final result will be, but here some attempts to show what works and doesn’t work.
• The third way is the Field of Dreams method.  This is specifically a design research method, and is suitable for either explanatory or exploratory research.  In your first study you build a tool, and then in subsequent studies use that tool to explore a community, dataset or phenomenon.   This differs from the Yellow Brick Road method in that you don’t necessarily know where you’re going with the tool, or what your final research question will be.  Your dissertation might end up looking like a Yellow Brick road dissertation, because after the fact, you an go back and tell a clean story out of what you found.   Fernanda Viegas and Mark Wattenberg are a good model for this in their work with manyeyes.org (Viegas et. al. 2007b, Viegas et. al. 2008) and  HistoryFlow (Viegas et. al. 2004, Viegas et. al. 2007a), as is Marc Smith with Netscan (Smith 2002, Burkhalter and Smith 2004) .  Although none of these examples are dissertation work,  I don’t see why you couldn’t use the method for one.  It is a bit of a gamble.

I had planned to use the Field of Dreams method, but now I’m starting to suspect that I don’t know enough to write the tool just yet.   So perhaps I’ll end up doing a modified  Yellow Brick Road method: do enough studies to learn how to build the tool.

References

• Burkhalter, B. and Smith, M. (2004). Inhabitant’s uses and reactions to Usenet social accounting data. Springer.
• Smith, M. (2002). Tools for navigating large social cyberspaces. Commun. ACM, 45(4):51-55.
• Viegas, F., Wattenberg, M., and Kushal, D. (2004). Studying cooperation and conflict between authors with history flow visualization. In Proceedings of the 2004 conference on Human factors in computing systems, pages 575-582, New York. ACM.
• Viegas, F. (2005). Revealing individual and collective pasts: Visualizations of online social archives. PhD thesis, Massachusetts Institute of Technology, School of Architecture and Planning, Program in Media Arts and Sciences.
• Viegas, F. B., Wattenberg, M., Kriss, J., and van Ham, F. (2007a). Talk before you type: Coordination in wikipedia. In System Sciences, 2007. HICSS 2007. 40th Annual Hawaii International Conference on, page 78.
• Viegas, F. B., Wattenberg, M., van Ham, F., Kriss, J., and McKeon, M. (2007b). ManyEyes: a site for visualization at internet scale. IEEE Trans Vis Comput Graph, 13(6):1121-1128.
• Viegas, F. B., Wattenberg, M., McKeon, M., Ham, F., and Kriss, J. (2008). Harry potter and the meat-filled freezer: A case study of spontaneous usage of visualization tools. In Hawaii International Conference on System Sciences, Proceedings of the 41st Annual, page 159.