An innovative new seminar titled Cyberinfrastructure in Science was taught this spring at the LTER Network Office (LNO) at the University of New Mexico (UNM). The virtual seminar was funded by the National Science Foundation's Office of Cyberinfrastructure and taught at three institutions in New Mexico and Arizona using videoconferencing technologies. The three institutions UNM and the Universities of Arizona (UA) and Northern Arizona (NAU) offered the course for credit to graduate students, but a broader group of faculty and researchers also participated just to learn the information.
According to Deana Pennington, the organizer and Principal Investigator of the seminar, together with Mark Servilla and Barney Maccabe as Co-PIs, the project arose out of the realization within the scientific community that problems that need to be tackled in the coming decades cannot be tackled by one person or discipline. "Problems such as global change, environmental change, interactions between environments and society, and how people understand the environment are interdisciplinary. And they really require inputs from numbers of scientists to try to come up with good scientific explanations," Deana says.
Deana says it is very difficult for scientists to work with their technology counterparts because the two groups have completely different perspectives that distract from their research problems and make it difficult for them to collaborate. "One of the things that became apparent early on was that there are social, cognitive, and technical aspects to collaboration," she says. "The technical problem is because the scientific data and understanding are spread out across different kinds of sciences around the globe; there are different data stores and ways of archiving information, so if we want to conduct collaborative science, we have to bring together their data and observations and that's a technical problem. The social and cognitive problems are more difficult to address. We have to enable rapid learning about selected, relevant parts of our collaborators' conceptual framework in order to associate those with our own knowledge and understanding."
Deana came up with the idea for a model of collaborative learning that she submitted to NSF as a Cyberinfrastructure team (CI-TEAM) project. The project was approved and funded for one year to demonstrate proof-of-concept. She points out that the approach could be replicated by any group of scientists interested in incorporating advanced technology into their research, adding that the project's goal is to implement a process model of collaborative learning that solves both the social and cognitive aspects of conducting interdisciplinary science. "You can't just place a group of researchers from computer science and multiple other sciences in a room and expect them to collaborate and come up with something useful," she cautions. "There has to be prep work -- a person has to be ready to collaborate."
According to Deana, the challenges include not only how to enable such collaboration, considering the diversity and spread of the disciplines involved, but also how to develop methodologies that make collaboration easier. While calls for interdisciplinary science have been common for a while, she says doing that has turned out to be much more difficult than anybody anticipated. "There is a real need to understand collaboration processes and theories. How can we make this happen? How can we make it easier? Are there ways and approaches that we can use to try to enable this process?"
To better tackle this challenge, Deana had to do a little bit of learning herself. She enrolled in classes in UNM's College of Education that specifically target adult learning in group settings, focusing on teamwork, group processes, and learning on the job. Furthermore, she realized that the business arena was very actively involved in promoting adult learning, so she brought that perspective to bear on the problem at hand getting scientists and computer scientists to work together collaboratively.
Her approach uses constructivist theory from the field of education, which holds that people construct mental models in their heads of what they know, and the way they learn is to add on to those mental models systematically. "That works really well as long as what you're trying to add on is fairly close to what you already know," she notes. "But when you are trying to add on something completely new, you're starting from scratch and there's a high cognitive load. There are certain things called intrinsic load that you really can't reduce -- there is a certain amount of information load that you really can't do anything about. But you can assist the cognitive construction process by chunking information, creating conceptual blocks, and systematically adding that information.
The subject matter of the seminar emphasized conceptual understanding of what technical approaches are being developed, why, and how those will enable future science. Scientists, computer scientists, and software developers from around the country who are currently working on next generation technologies for biologists, ecologists and environmental scientists taught the seminar remotely. She says, "I was able to get some of the leaders in the cyberinfrastructure arena to do remote presentations because it only takes a few hours of their time. Flying them in to do a traditional seminar wouldn't be workable because most of them don't have the time to do that."
The classes included demonstrations of ongoing work, and some of the technologies under the hood were discussed at a high level, emphasizing important technical concepts rather than technical details. The seminar used Adobe's Breeze system, which allows a limited number of people to connect to each other by live video and audio and for presenters to share their desktops with participants to demonstrate software off the desktop. While it might not be as good as having everyone in the same room, Deana says it is the only way to do it if people are distributed. Participants got some experience just by being in the seminar, but ultimately, "you want to be able to do science, conduct analysis, run models and engage in all sorts of activities in real time while people are distributed and at a distance," she adds.
Deana views it all as a series of nested experiments. "The first experiment is that I have hypothesized a model of what I think will make it easier for cross-disciplinary scientists to work together, but we don't really know ... nobody knows how to do that as yet. Even deeper than that is an experiment of what technologies enable computer scientists to enable cross-disciplinary science. So there are many levels of testing that we're doing."
The seminar ended in late April 2007. Subsequently, the group held a working meeting to collectively decide how to proceed and try to come up with some unique and innovative science. Ultimately, Deana says, they will write a proposal for more funding to do the actual work. Meanwhile, recordings and pdfs from the sessions are posted online and can be downloaded at www.scidesign.org.