Future CSD Workshop

Today I participated in the CSCW 2012 workshop on the Future of Collaborative Software Development (Future CSD). This was a visioning workshop on important transitions in the world of software development and implications for collaboration technology.

The workshop organized around five themes, each kicked off by a presentation followed by a discussant and group discussion. Participants simultaneously created a collective google doc for each theme while presentations were going on. As a side note, these docs turned into a great summary of the workshop as well as an interesting backchannel for deeper discussion of the ideas being presented.

We presented and discussed our research on social transparency and software development practices, under the theme of Increased Awareness:

Laura Dabbish, Colleen Stuart, Jason Tsay and Jim Herbsleb. Coding for an Audience: Transparency and Collaborative Behavior in a Social Coding Environment.

Jennifer Marlow and Laura Dabbish. Understanding interpersonal mental model formation in distributed software development: A case study in GitHub.

Jason Tsay, Laura Dabbish and James Herbsleb. Work Concentration and Success in Decentralized Teams.

It was great to hear about other people’s work in this space and discuss what the future holds for collaborative software development! There were some interesting (and sometimes conflicting ideas) about how the world of software development is evolving and how collaboration support should evolve along with it.

CSCW 2012

I’m headed to CSCW 2012 (the ACM Conference on Computer-Supported Cooperative Work) in Bellevue, WA with my co-PIs and students.

We’ll be presenting three papers at the conference all related to social transparency in online settings. We define social transparency as the visibility of information about other’s behavior online, including their identity (who is it?), actions on artifacts (what are they doing or looking at?), interactions with others (and who are they talking to?). In our recent research we have been theorizing about this concept, examining how it influences collaboration and impression formation in a naturalistic way, and conducting experiments manipulating features of the transparent signal and examining the influence on participation in a social setting.

Our first paper is on Monday Feb 13th at 11:00 (in Grand C) in the session on Community and Classification Online. This is a note that was also selected as a Best Note in the conference!:

Fresh Faces in the Crowd: Turnover, Identity, and Commitment in Online Groups
Laura Dabbish, Rosta Farzan, Robert Kraut, Tom Postmes

This note presents an experiment on the influence of visible turnover and a common social identity on member retention in online social groups. Here visible turnover was manipulated in terms of the amount of change in member faces individuals are exposed to in an online group. Interestingly we found that turnover had distinct effects on member participation and retention depending on the presence of a common social identity (enhancing rather than depressing participation with more salient indicators of a common identity).

Our next paper on Monday Feb 13th at 3:00 (in Grand A) in the session on Twitter and Social Transparency:

Social Transparency in Networked Information Exchange: A Theoretical Framework
Colleen Stuart, Laura Dabbish, Sara Kiesler, Peter Kinnaird, Ruogu Kang

In this paper we delineate the concept of social transparency and introduce a framework for theorizing about how different types of social transparency may influence collaborative outcomes such as knowledge transfer, team performance, innovation, and relationship development.

Our third paper is on Tuesday Feb 14th at 11:00 (in Grand J) in the session on Toolkits and Software Development:

Social Coding in GitHub: Transparency and Collaboration in an Open Software Repository
Laura Dabbish, Colleen Stuart, Jason Tsay, Jim Herbsleb

This paper presents a qualitative study on collaborative behavior in a transparent work environment. In our study, we looked at how the social transparency afforded by social media functionality in the site GitHub helps software developers find new knowledge and coordinate their work. We wanted to understand social transparency in action in an large online community, particularly how people interpreted cues about others’ behaviors online and how those inferences affected work practices.

My collaborators on these papers have been excellent and thought-provoking! In all we are presenting three very different kinds of pieces (conceptual, exploratory qualitative, and experimental quantitative) advancing our understanding of social transparency in online environments.

Constructing Association Networks

We have an article coming out in this month’s special issue of Statistical Analysis and Data Mining on ‘Networks':

Dabbish, L., Towne, B., Diesner, J., & Herbsleb, J. (2011). Construction of association networks from communication in teams working on complex projects. Statistical Analysis and Data Mining, 4(5), 547-563. [Access Online Here]

In this paper we present a presenting a technique for extracting associations among components in complex project work using the contents of member’s communication. We believe these kinds of association networks can be used to identify dependencies among project components and channel communication and information to project members.

SCALE Project Kickoff

This week was our kickoff meeting for the SCALE project, an NSF-funded research collaboration on ultra-large scale software development. In this project we are working to better understand and develop technologies to support collaboration and coordination needs in very large projects that span organizational and geographic boundaries. Learn more on the project website here: http://www.coaste.org/scaleresearchgroup/

During our kickoff meeting, project members from Carnegie Mellon University, University of California at Irvine, University of Nebraska, and Bosch all convened in Pittsburgh to discuss project goals, present our preliminary results, and brainstorm on next steps. The research portfolio of our team represents the ideal combination of empirical work on organizational psychology (Brandy Aven and Linda Argote) collaboration and technology use (Laura Dabbish and Jim Herbsleb), and design and evaluation of innovative software development coordination technology (Andre van der Hoek, David Redmiles and Anita Sarma). In our sessions we identified interesting ways our research complements each other and identified collaboration opportunities within the team. Ultimately our goal is to develop coordination systems and organizational processes that will make large-scale and distributed project awareness, communication, and coordination as effortless as in a small team. We are off to a great start and I’m really looking forward to working together!

NSF VOSS: Building Social Attachment in Virtual Groups

The existence and survival of online virtual organizations depends upon the commitment and retention of their members.

I am the Principal Investigator of an NSF VOSS grant on ‘Building Social Attachment in Virtual Groups’ (Award #0943159), with co-PIs Robert Kraut (Carnegie Mellon University) and Tom Postmes (University of Groningen) investigating alternative ways of designing online sites to increase commitment to virtual organizations based on social science theories of attachment. In a series of experiments within different virtual settings we are varying features of the groups that social science theory predicts will promote different forms of attachment and examining their influence on member participation and retention. Our results help inform the design of virtual groups and online communities to foster member commitment.

Check out our project website where you can find more information about the project goals, team, and publications.

INGROUP 2011

This summer I presented our ongoing research on designing for commitment to online groups at the sixth annual conference of the Interdisciplinary Network for Group Research (INGRoup 2011), July 21-23rd in Minneapolis, MN:

Dabbish, L., Kraut, R., & Patton, J. Should I stay or should I go? Depends on how you talk to me: Communication and team commitment. Presentation at the Sixth Annual INGRoup Conference, July 21st-23rd in Minneapolis, MN.

In this work, we examined the role of communication in fostering commitment to a group. Previous work in organizational psychology has demonstrated a positive correlation between communication and organizational and team commitment. This previous work has largely been cross-sectional and survey-based. As a result, it remains unclear whether communication fosters commitment by getting people involved (being talked to leads to talking which ultimately creates commitment) or creating a social environment (being talked to is enough). We conducted a study in a massively multiplayer online game, where participants were randomly assigned to a group with a confederate. We varied the type of talk (silent, talked about on task things or talked about off task things) and the role of the person doing the talking (leader or peer). Our results suggest that both on task and off task communication fosters commitment by creating a certain environment (versus as a function of the participants own talk).

If you are interested in a copy of the working paper, please send me an email.

COASTE

Today is the kickoff meeting for COASTE, the Center on Architecting Socio-Technical Ecosystems, a planned Industry/University Collaborative Research Center sponsored by the National Science Foundation. The center will join leading researchers in socio-technical systems, software architecture, open source software, network science, and collaboration technologies to work with member organizations to rapidly uncover social, technical, and organizational principles to shape a new scientific perspective that will guide effective ecosystem design.