Collaboration in Emerging Ecological Practice
Human-Computer Interaction in
Biodiversity Informatics Workshop
June 2, 2005
Bonnie A. Nardi
Department of Informatics
School of Information and Computer Science
University of California, Irvine
My background
Anthropologist
Computer-supported collaborative work
Science studies (how does science work?)
Activity theory (culture and cognition)
Bonnie A. Nardi 2
Biodiversity Workshop, June 2, 2005
Research on collaborative tools for ecologists
Qui ckTime™ and a
TIFF (U ncompr essed) decompressor
are needed to see thi s pi cture.
Qui ckTime™ and a
TIFF (U ncompr essed) decompressor
are needed to see thi s pi cture.
Qui ckTime™ and a
TIFF (U ncompr essed) decompressor
are needed to see thi s pi cture.
Bonnie A. Nardi 3
Biodiversity Workshop, June 2, 2005
Why Ecology?
Pressing needs to understand environmental change
Moment in time: ecology transforming to big science
Activity theory asks how such transformations happen
Interesting issues of changes in collaborative
technologies and practices during such transformations
Bonnie A. Nardi 4
Biodiversity Workshop, June 2, 2005
General approach to my work
Ethnographic studies of work practices
Design of new tools
Prototype, test new tools to evaluate their usefulness and usability
working with technical people
Bonnie A. Nardi 5
Biodiversity Workshop, June 2, 2005
Research on “emerging practice” in ecology
Current practice: single investigator or small teams
Emerging practice: larger teams, more data, more technology
Transforming to “big science” (or trying to)
Bonnie A. Nardi 6
Biodiversity Workshop, June 2, 2005
“Big science”
Broad, problem-directed goals
Interdisciplinary teams of scientists, engineers, technicians
Work is often distributed
Extensive instrumentation (such as NEON sensor network)
Processing of large volumes of data
Projects are professionally managed
Bonnie A. Nardi 7
Biodiversity Workshop, June 2, 2005
Drivers of move to big science in ecology
Long Term Ecological Research Network (LTER)
EcoVisions Project of Ecological Society of America
NEON, CUAHSI (hydrology), CLEANER, and others
Availability of terabyte scale datasets from NASA, DOE, EPA,
USGS and other government agencies
Individual scientists discovering these datasets
Bonnie A. Nardi 8
Biodiversity Workshop, June 2, 2005
Collaboration in Ecology
As ecology scales up to big science, need for different
kinds of collaboration:
interdisciplinary
distributed
involves sharing and interpreting large volumes of data
What kinds of technologies will sustain such collaboration?
Are the problems similar to those of other fields?
Bonnie A. Nardi 9
Biodiversity Workshop, June 2, 2005
Something that might be unique to ecology:
Combination of:
Data from heterogeneous sources, including historical
field-based data which can be quite old
Very complex interactions among variables
living things in complex contexts
Bonnie A. Nardi 10
Biodiversity Workshop, June 2, 2005
What I’ve done so far
QuickTi me™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Bonnie A. Nardi 11
Biodiversity Workshop, June 2, 2005
Collaboration in Ecology Workshop at UC
Irvine, October, 2004
Karen Baker (data manager)
Stephen Bocking (historian who studies ecology)
Scott Collins (ecologist)
Paul Dourish (computer scientist)
Cliff Duke (ESA)
Anna Gold (librarian)
Bryan Heidorn (librarian)
Vivian Hutchison (USGS)
Roberta Lamb (computer scientist)
Renee Miller (computer scientist)
Gary Olson (psychologist)
Diane Pataki (ecologist)
Bonnie A. Nardi 12
Biodiversity Workshop, June 2, 2005
Collaboration in Ecology Workshop
Mark Schildhaeur (NCEAS)
Katie Suding (ecologist)
Bill Tomlinson (computer scientist)
Ferdinando Villa (ecologist)
Ann Zimmerman (librarian and co-organizer)
Bonnie A. Nardi 13
Biodiversity Workshop, June 2, 2005
What else…
NSF Grant submitted: Collaboration in Transformational Science
(with Ann Zimmerman at U Michigan and Susan Sim, UCI)
Ongoing interviews with ecologists at UC Irvine
Participation in NEON
Member NEON (National Ecological Observatory Network)
Design Consortium’s Information Technology and
Communication Subcommittee (“Cyberinfrastructure”)
Bonnie A. Nardi 14
Biodiversity Workshop, June 2, 2005
What else…
Ph.D. student will study seismologists at UC San Diego this
summer
How they work with technicians and engineers to configure
distributed sensors and design experiments
Bonnie A. Nardi 15
Biodiversity Workshop, June 2, 2005
Preliminary ideas on collaboration problems of
interest
“Dating”
Design
Data
Bonnie A. Nardi 16
Biodiversity Workshop, June 2, 2005
“Dating”
Problem:
Find collaborators with special expertise in a
particular area of ecology or in another field
As in dating, half the problem is locating the right person and the
other half is getting the relationship to work
Possible Solution:
Online directory (like Friendster)
Service for linking people to collaborate in small
light-weight projects to get to know each other
Bonnie A. Nardi 17
Biodiversity Workshop, June 2, 2005
Design
Problem:
Experiments with large sensor network
Or analyses with government databases
How to design large experiments, analyses, observations?
Solutions: ??
Need ethnographic work here
Bonnie A. Nardi 18
Biodiversity Workshop, June 2, 2005
Data
Problem:
NCEAS et al. working on data problems of data
integration, metadata
How to interpet data scientists did not collect,
especially when there are large volumes of data
Bonnie A. Nardi 19
Biodiversity Workshop, June 2, 2005
Interpreting data
“Data are not simple carriers of meaning. [C]onverting raw data
into scientific or social meaning is an active, context-dependent
process” (Birnholtz and Bietz, 2003)
Solution:
Talk to others about their experiences with the data
Bonnie A. Nardi 20
Biodiversity Workshop, June 2, 2005
People talk about data to interpret it
Zimmerman interviewed an ecologist who was formerly an
economist:
In economics, typically people are working with a shared data set.
There are hundreds of people that work with the current population
survey, for example, and you can go and find out, “Well, what are
the problems with this data set?” Everyone can tell you, “Oh yeah,
’79 was a really bad year, and there’s a glitch, and you are going to
have to reprocess this field if you want to use it.”
Bonnie A. Nardi 21
Biodiversity Workshop, June 2, 2005
Solution: Data Conversations
Communications research has shown that conversation is
efficient for clarification, coordination, updates, brainstorming,
interpretation, critiquing and elaboration of complex ideas.
Harvest electronic conversations in instant messaging, wikis,
blogs, chats, listservs.
Make available to wider group of scientists
Tools for summary and presentation needed
Bonnie A. Nardi 22
Biodiversity Workshop, June 2, 2005
Conclusion: Outputs of research
a. New understandings of transformational science
b. New collaborative tools for ecologists working toward
transformational science
c. Testing of ideas about collaboration in activity theory
(which involve a and b).
Qui ckTime™ and a
TIFF (U ncompr essed) decompressor
are needed to see thi s pi cture.
Bonnie A. Nardi 23
Biodiversity Workshop, June 2, 2005