Seattle, WA
Facilitator: Don Spicer, University of Maryland System
Facilitator: David Stack, University of Wisconsin–Milwaukee
Guest Presenter: Mark Sheehan, EDUCAUSE
Mark Sheehan is the PI for an upcoming ECAR mini-study on cyberinfrastructure that was commissioned because there were gaps in the previous ECAR study of IT Engagement in Research, e.g., external funding and resources.
Net@EDU has held several meetings on cyberinfrastructure. The July 2007 summit had several requests for action from EDUCAUSE, including the ECAR mini-study. Net@EDU is looking for closure on certain issues and validation of the work of the 50 or so people who took part in their workshops.
It was pointed out that the not all researchers are created equal, and a study will need to validate whether or not the "mountain lion" researchers make use of cyberinfrastructure. ECAR may request that both CIOs and VPs for research, or similar high level stakeholders, complete the survey together to provide one response that describes the entire institution, not just central IT.
People from some institutions expressed the desire to benchmark against peer schools/colleges. ECAR has a study of research in medical schools that is about to be released.
The proposed study will be a matrix of tracks:
versus estimations of:
It was recommended that additional tracks of software licensing and governance structure be added but, unless something is removed, each would result in 24 additional survey items which would unduly complicate the analysis.
Likert scales of the respondents' perceptions will be used rather than measures of flops, dollars or percent utilizations. It was questioned whether institutional responses might tend to result in average values across many of the responses. The survey may identify areas for future studies to drill into.
Each section of the study may have qualitative questions as well. ECAR usually does qualitative follow-ups with specific institutions.
The activity levels would be difficult and time consuming to accurately quantify, e.g., institutional teraflops. If too much effort is required, people will abandon the survey and the response rate may be too low to yield valid results.
By asking a combination of questions regarding importance and effectiveness, it will be possible to get a qualified semblance of the truth. The investments of human resources in research support recorded in the previous ECAR survey are one measure of importance. Another would be to ask CIOs, VPs for research, provosts, etc., how they feel their cyberinfrastructure efforts compare in importance to those of their peer institutions.
So far, not many people seem to be saying that cyberinfrastructure is important for teaching and learning. On the other hand, some say that rather commonplace technologies, such as learning management systems, are cyberinfrastructure. For others, the concept of cyberinfrastructure implies staying ahead of the curve. Perhaps the relationship of cyberinfrastructure to teaching and learning should be explored.
ID management is one of the most challenging pieces of middleware and InCommon will be needed to access NIH resources. Research groups are starting to see the need for ID management expertise from central IT organizations.
The mini-study will need responses from the entire spectrum of institutions because the needs and challenges of cyberinfrastructure are likely to be different up and down the chain. The results will likely need to be profiled by different types of institutions, e.g., Carnegie classification.
Although actual institutional numeric data would be difficult to get, it is valuable to ask for gut level responses on all sources of funding for research IT, both internal and external.
It would be interesting to explore whether there are any governance structures related to resource allocation, recharging and whether grant overhead funds are being used to support cyberinfrastructure.
Another possibility would be to inquire regarding an institution's priorities for where additional funding should go, e.g., discretionary raises, cyberinfrastructure, tuition reduction, etc.
There are cyberinfrastructure challenges other than technology, e.g., managing grant finances, copyright, intellectual property and lack of adequate data center resources. ECAR may get to studying these issues. Perhaps there should be an open-ended question about what the survey has not addressed.
The group also discussed whether the name of the Research Mission Support Constituent Group should be changed to Cyberinfrastructure and Research Support. In a context broader than research, cyberinfrastructure can be seen as a construct for the institution in terms of planning and engagement on strategic goals. Cyberinfrastructure may also provide a mechanism for re-engagement of central IT with PIs.
Institutions have lost grants due to a lack of a central cyberinfrastructure to support the research enterprise. As funding agencies look for multi-institution and multi-disciplinary participation in grants, they will drive the need for cyberinfrastructure. Unfortunately, current communication and governance structures tend to be insufficient, fractured or uncoordinated.
On another topic, the EDUCAUSE Current Issues Committee is looking to incorporate cyberinfrastructure into its annual survey of top IT issues. The two most likely choices are either a subtopic under infrastructure or a renamed main topic research support main topic. Despite the earlier discussion about the purview of cyberinfrastructure extending beyond research, there was consensus that for 2007 cyberinfrastructure should be incorporated in the name of the research support topic.