:::: MENU ::::
Browsing posts in: Collaboration

Not my dissertation but close to my heart

Here’s a summary of the other research I do (or want to do) that’s not my dissertation. This was originally a 2-page research statement submitted to some people with money to burn.

Research Summary – Communities and Technologies

At the broadest level, my research is about communities and technology. My research enriches our understanding of the roles social media play in supporting offline communities. My approach differs from much of current social media research, because I focus on teams and organizations in which people know each other and use technologies to support their activities (e.g. Upcoming!, workplace wikis) rather than on online communities of people who do not know one another offline (e.g. SlashDot, Yahoo! Answers) (see Beenen et al., 2004; Lampe & Resnick, 2004; Preece, 2000; and Smith & Kollock, 1999). My work necessarily encompasses studying offline behavior as well online behavior; in order to understand social media use by groups, it helps to understand the nature of a community. My research highlights the situated nature of social media use by offline communities and focuses on how social and technical processes impact community behavior both online and off. A better understanding of behavior in communities using social media enables us to design social media more effectively and to recommend behaviors and tools to make communities more successful.

My most recent work asks, how do faculty and students in a graduate school use a wiki to share information about their community with each other and with the public? What does their use tell us about what it might be important for new community members to learn? How can we use their wiki use behavior to understand how people make decisions about what information to share and what to keep to themselves? Understanding the community provides insights into the way members of those communities interact with one another via social media. My goal is to leverage human and computing resources so that a sociotechnical system can use the skills of humans and benefits of computation to improve collaboration and its supporting technologies. The remainder of this document briefly describes projects in which I have been involved with and ends with an overview of my continuing work.

Sharing and Storing Community Knowledge

In an era when more than half of all doctoral students leave before finishing their degrees and students must compete for increasingly scarce human and financial resources, it’s no surprise that students welcome help completing their degree requirements. What is surprising in this instance is that students are not just the primary consumers of the information but are also the primary producers. They share human subjects review applications, books that help them write dissertation proposals, interview protocols, even advice about how to set up an experiment using existing technical resources. We might expect students competing for the same pool of resources to hoard, but in this instance, students are much more collaborative than competitive. Their behavior on the wiki demonstrates this difference, but only by studying the offline community can we really understand why. In this case, it’s likely that the collaborative ethic of the school itself permeates the doctoral students. Faculty and students at the school, regardless of whether they use the wiki, recognize and enjoy the collegial atmosphere of the school. Students are well-funded by research and teaching positions and are encouraged by their faculty’s examples and instructions to work together to do better research. The wiki is not the reason students share, but it is the social media tool they use to do so.

Another aspect of the wiki example that I find interesting is the near-mashup nature of content created and the potential such behavior indicates. On the wiki, users include data available elsewhere but combine that data in community-specific ways. For instance, one wiki page serves as a marketplace for used textbooks required by courses within the school. That page includes data from the course syllabi, email lists, booksellers, and individual users. Such pages indicate community information needs – in this case, students need to sell their extra books to a small potential market while students in that buying market seek good deals on books and some advance warning of what textbooks they’ll need. Such pages also indicate what potentially useful mashups might appear were users able to construct them. New social media that offer and use open APIs such as Yahoo! Pipes, Yahoo! Maps, and Upcoming! make asking such questions – what data sources might users combine for their communities if they could do it themselves? – possible.

Facilitating Ad Hoc Ridesharing

I was part of the original RideNow team at the University of Michigan. Our goal was to facilitate ad hoc ridesharing in Ann Arbor and to develop technologies that could be used to do the same in other communities. Cars in the U.S. can comfortably seat four or five people but rarely carry more than one (Transportation Statistics, 2004). Filling some of those seats would create tremendous benefits for both individuals and society as a whole. Riders and drivers would have convenient travel and the possibility of pleasant conversation. Society would benefit from reduced emissions and road congestion. However, barriers to ridesharing include 1) coordination problems, 2) risks of riding with strangers, and 3) mismatch in cost and benefit for riders and drivers.

We designed a service, called RideNow, that approached the problem of ridesharing by capitalizing on the benefits of incremental and localized design. Our system avoids the costs of overengineering by allowing incremental changes to occur. For example, the first instance of the system was rather bare bones – it offered free text fields that allowed users to decide how to specify ride information. Later versions of the system offered structured fields based on the behavior users exhibited in the first system. For example, the second generation of RideNow can parse dates such as “next Friday” rather than requiring a user to enter a specific date. The system also capitalizes on the benefits of nuance and ambiguity afforded by localized design. For example, RideNow’s data fields allow users to enter information such as “after the faculty meeting.” Our goal with RideNow was to design a system that allowed a well-established community to use personalized, situated software (Shirky, 2004) and that remained flexible enough to be adopted by other communities.

Continuing Research

My future work will extend my interest in studying communities and designing/building software to facilitate their collaborative activities. It is important to me to have a close connection between field research and system design. As social computing tools become more prevalent and the distance between developer and user diminishes, opportunities to improve both development and use abound. I look forward to asking question such as, how can we make powerful mashup tools such as Yahoo! Pipes usable by non-developers? What would users do with such technologies if they could use them? How would users tailor the content of their mashups and contributions to specific community audiences? I have seen users embrace flexible, situated technologies such as ridesharing systems and wikis, and I believe there is great promise for end-user development of social computing technologies. Issues such as community building and information sharing generalize regardless of the community being studied, and I look forward to the opportunity to study social computing and larger, distributed offline communities such as political movements and distributed work teams.


(2004). Omnibus survey household survey results. U.S. Department of Transportation Bureau of Transportation Statistics.

Beenen, G., Ling, K., Wang, X., Chang, K., Frankowski, D., Resnick, P. & Kraut, R. (2004). Using social
psychology to motivate contributions to online communities. In Proceedings of the Conference on
Human Factors in Computing Systems CHI 2004

Lampe, C. & Resnick, P. (2004). Slash(dot) and burn: Distributed moderation in a large online conversation space. In Proceedings of the Conference on Human Factors in Computing Systems CHI 2004 (pp. 542—550). Vienna, Austria: ACM Press.

Preece, J. (2000). Online Communities: Designing Usability and Supporting Socialbility. New York: Wiley.

Shirky, C. (2004). Situated software.

Smith, M. & Kollock, P. (Eds.). (1999). Communities in Cyberspace. Routledge.

New Media consulting

Yes, I am available for hire. I’ve received some email inquiries about whether I’m available for contract work. I am indeed. If you’re looking for help

  • Setting up a wiki, blog, or intranet,
  • Analyzing wiki or blog traffic and use,
  • Writing surveys and interview protocols, or
  • Analyzing survey and interview data,

send me email. My recent work focuses on improving collaborations and facilitating organizational knowledge sharing. I also have 8 years of web development experience (ASP, SQL Server, PHP, MySQL, JavaScript) and 5 years of research experience (interviews, surveys, statistical analysis). I have installed blogs and wikis in a variety of organizations and environments and have run workshops for individuals interested in setting up their own. I can give workshops to large and small groups about what new media technologies are, how they can be used, and what pitfalls they present.

Thank you for your interest!

Solving the service provider problem

During the CDI Workshop at Rennselaer in September, one of the computer scientists complained that when he collaborates with social scientists, he feels like they view him as a service provider rather than a collaborator. It sounded like he had some experience with a social scientist who’s approach was to say, “Go build this thing so I can deploy it and study the deployment.” In my short talk, I mentioned that social scientists in interdisciplinary collaborations are not service providers either. I’ve worked with computer scientists before who approach our work with the attitude that I will “fix the social stuff”, whatever that means. So it seems that we have a problem. Computer scientists and social scientists recognize that if we worked together, we might find answers to interesting problems. Here I’m thinking about expertise finding, knowledge sharing, and distributed collaboration as problems that might benefit from such a collaboration. How can we work together without having either side feel like the other side is using them for a service rather than as a colleague?

Man, I wish I had an answer. Why is this bothering me today? Well, I’m trying to set up CoSign so that the new version of the KNOW SI wiki will allow UM users to login using their existing UM login credentials. This means I need to dig into the innards of the Apache server we’re running. That sounds almost CS-y to me. Probably not to a CS person though. Anyway, CoSign and the resulting permissions options represent one of the socio-technical problems that I think could benefit from both computer science and social science. What’s the best way to set up permissions on our school wiki? What technical infrastructure (e.g. .htaccess, CoSign, MediaWiki extensions) is necessary to supporting the kind of social behavior (e.g. TBD, which makes the technical questions that much harder) in which people want to engage on this wiki? Those are the questions I’ll be wrestling with this weekend and probably for a while.

Explaining myself to technology product groups

I have a meeting tomorrow with a leader of a product group at a Silicon Valley-based household name. On the itinerary, I’m described as “a PhD student who has worked on use of social media and etc. by science and engineering research groups”. I’ve been asked to come up with a 3-5 minute introduction for myself. Here’s my rough draft:

Your itinerary shows that I’ve worked on social media use by science and engineering communities. Another way to think about what I do is to call it social computing for existing (or emerging) social systems. My work is with distributed teams, and it’s different from much of the existing ecommunity work because the teams I study know each other. They may not be close socially or geographically, but they are familiar with each other and have almost always met face-to-face. They’re working together toward some goal that requires them to collaborate (e.g. getting a new building technology to market) or benefits greatly from collaboration (e.g. getting through grad school). They span a variety of distributions — whether their offices are states or continents apart, whether their disciplines seem distant, and how they span the spectrum from novice to expert.

Two of the projects I work with now that explore the use of social computing by existing communities are the CI TEAM grant and KNOW SI. In the CI TEAM grant we explore how engineers developing and testing a new building material work together across institutional and national boundaries to establish standards for testing the material and for training new users of the material. The CI TEAM group uses a content management system with features somewhat like Yahoo! Groups to share their data and discuss it. CI TEAM builds on earlier large-scale scientific collaborations that used the same tools but had very different motivations and goals for collaboration. What I’ve found most interesting in these scientific research collaboration projects is how unlikely users are to adopt a technology specifically for their projects. Users love email and Excel but ignore wikis, archived email lists, and file repositories.

KNOW SI is a PhD student-led project with many goals. Mine is to use KNOW SI as a way to understand how people in organizations, such as schools, use available technologies to share information with one another. In KNOW SI, I helped set up an iterative series of wikis for use by the community, and we’re analyzing that use and preparing the next generation wiki. What’s surprised me here is how willing various groups are to use the same underlying technology. For example, doctoral students are all over the wiki. We’re our own audience though. The Research and Career Services offices have different audiences — master’s students and the public for two — but are exploring the same technology.

Together these projects provide a variety of settings for me to explore information and technology use in different kinds of collaborations.

Transformation of practices

Here’s some text from a recent proposal draft (I’ll add citations later). I’m still editing the document from which these are excerpts, but the people have spoken, and they asked for drafts. So here you go. This is what I’m working on now.

Much of the research on practice focuses on how it may be transfered from one firm to another, from one
person to another, or from one group within an organization to another part of the same organization.
This study builds on those literatures, but asks a different question – how do networked practices change when
one or more parts of the network change? Instead of exploring a sender-receiver model of transfer of practice,
this study explores scenarios in which the source and target of a new practice are the same. Instead of focusing
on practice within a single community of practice (CoP), this study explores how multiple, interacting CoPs
influence one another. To refer to the set of changes that occur in the network, I use the term transformation
of practices. The following describes relevant terms and literature and proposes a study designed to produce
data necessary to describe the processes involved in transformation of practices. The plural of ”practices” is
necessary here; that I explore the relationships among communities of practice and their impact on one another
sets this study apart. The goals of the study are to describe the network of actors in such a way that enables us to understand the
practices in which those actors engage and how those practices relate.

The transformation of practice seems like a learning and coordination problem. First, someone must develop a
new material or method – broadly a new technology – that is a candidate for adoption by the network. Then, the
technology must be successfully adopted by a number of communities within the network. This sounds much
like Rogers’ diffusion of innovation work, but there still he described the uptake of innovations by people
engaged in the same kind of work. Here, the problem is a little different in that many communities are pursuing
a common goal, a technology with the potential to change how that goal is achieved is introduced, and each
of those already distinct practices must adjust to account for the new technology. This proposal describes a
study that focuses on a case of a transformative technology – engineered cementitious composite (ECC) – and
the resulting transformation of practices within the civil infrastructure building network.

I want to be able to talk about something like a network of practice (NoP). Brown and Duguid characterize
NoPs as members sharing a common practice but not needing to coordinate their work. I’d rather think of
an NoP as members needed to coordinate work but whose practices are not the same. The members have a
common goal (e.g. build a bridge) but none of them do the same thing (e.g. design bridge vs. pour concrete).
This kind of activity seems more networked to me than Brown and Duguid’s characterization. However, I don’t
want to use NoP if a big name already did and means something different from what I mean. What else could
I call it? I’m thinking of practice at a higher level of granularity, maybe? Maybe I mean ”system” and not

Phenomenologist = me?

I had an interesting meeting with Gina Venolia this morning during which she used the term “phenomenologist.” I haven’t heard that term in a while, but it was a welcome utterance for sure. Gina and I were talking about knowledge – how it is used in teams, how it moves among people, how it gets captured and embellished in boundary objects. I was somewhat surprised to have an 80 minute conversation with a Microsoft Researcher and not have the topic of software developer (as a researcher’s objective, not as an area of study) come up. I have tremendous respect for much of the work in which Microsoft Research people I know engage, but I’ve always read it as having an eye to what Microsoft might develop for sale next. Not in a bad way, just in a way that’s very different from the product agnostic approach of academic research. Gina seemed welcoming of my phenomenologist tendencies (to study phenomenon with an eye to describing them) and unphased by my explanation that I no longer spend much energy thinking about systems I could build. The idea that I wouldn’t have to spend all my energy thinking about how to build a technical system using the knowledge I learned studying a phenomenon makes me more excited about doing just that. Funny how reverse psychology (or something like it) works, eh?

Gina’s work with and about software development teams and their mental models of code and my work with vaccine developers and civil engineers had a number of remarkable similarities. We both had stories to share about the development and use of boundary objects and how they require embellishment by a human in order to be useful. We both focused on the knowledge activities in these domains – those activities where knowledge is used, shared, clarified, developed, transformed.

Those 80 minutes were another welcome occasion for me to talk about my work with someone outside my lab, and that activity always serves to help me refocus and refuel. It’s easy to forget why something is interesting or significant when one is in the thick of it, but such meetings provide opportunities to talk about the forest through the trees. Sigh. Lovely.

From May – a found post (networks, communities, practice)

Here’s a little blog post from a couple months ago that hadn’t made it off my laptop and into the world. I’ve edited it a bit, but most of the text is from May. I was reconsidering the communities of practice literature.

The project I’m currently funded on is ostensibly about facilitating the implementation of civil infrastructure. When I was first presented with the project, the part that seemed most interesting to me was the “transfer of practice” (TOP) problem. The TOP problem goes something like “it’s difficult to move practice from research labs to the real world.” Sure is. I looked forward to working on that problem. As I got into the project more, my focus changed. It seems like now the problem is not so much how do we move a practice from over here in research land to over there in construction but rather, how do the practices of civil infrastructure design and construction change when the materials available change?

Engineered cementitious composites (ECC) have the potential to change the practices of civil infrastructure design and construction. I don’t know enough about that design and construction to yet know what the possibilities are, but I get the sense that they are big and dramatic. Iron and steel certainly made a big difference. Concrete, the rigid kind, is sure important. Imagine what happens when you change the tools again! At least, that’s what I’m imagining. With a little help from my colleagues, I’ll do some more definitive imagining.

So what does any of this have to do with communities of practice (COPs)? The problem of TOP is something like moving a practice from one community to new individuals. Here, I’m describing what happens when new people learn about ECC and start to learn how to work with it. It’s tough to make — the recipe is incredibly precise and the underlying theory is important — and it’s deceptively similar to regular concrete. To solve TOP, you simply send that newly trained person off in to the world, much like a PhD from the lab at Michigan is now off in the world getting his company to use ECC. Obviously I’m oversimplifying here, but you get the idea. By characterizing the problem of developing new infrastructure as a TOP problem, we make the research lab and its practices the goal, and the “real world” and its practices the target. This could even be a transfer practice from one community to another problem.

However, that’s not what I think is going on. Rather, I think the communities of practice involved are a little broader than that TOP conception allows. I don’t think the problem is one of trying to ease the problems with throwing ECC over the wall between lab and field. The really interesting problem, I think, is how does a change in practice within the civil infrastructure design and construction community happen? Much of the existing COP literature is about moving practice from a community to an individual. What about changing a practice within the community? How does that happen?

To start to answer those questions, I’m off to explore a variety of literatures including innovation (generally), innovation (in construction), organizational change, apprenticeship, public policy and infrastructure, and standards development and negotiation. That’s just to get me started. I’m likely to blog about this quite a bit in the near future as I try to figure out what I think is going on, or rather where I think something interesting is going on. In doing so, I hope to avoid the plague of reductionism against which Latour warns. Instead, I’m looking for the details of the network of forces at work in these communities and affecting their practices.