Analysis of Data
from the RCAC Principal Investigator Information Technology Survey
Conducted by the Research Computing Advisory Committee
with the Office of Research
Fall 2001 and Winter 2002
Executive Summary
During fall quarter 2001 and winter quarter
2002, The Research Computing Advisory Committee worked with the
Office of Research to conduct a survey of faculty researchers who
had served as principal investigators (PIs) on sponsored research
proposals, whether funded or not. The survey was designed to
assess the active PI perspective on information technology support
at OSU, the future direction of information technology for
research, and how computing technology at the university aided or
hindered their capacity to secure grant funding. The survey was
mailed from the Office of Research to active PIs of record. While
only about 35 PIs responded, the qualitative data presented at
least a useful anecdotal picture of faculty concerns from a
motivated group. The PlanIT Current State Assessment team
re-analyzed the source data for the planning initiative.
Survey questions were open-ended and a content
analysis of responses was conducted. Given the diversity of
specialty areas within the sample, many of the responses were
researcher-specific. The primary concern of PIs was university
funding levels for both equipment and technical support. Beyond
funding and support issues, PIs are most interested ease of
access, computing capacity and speed, parallel and
high-performance computing, real-time processing, and the
consistent upgrading of equipment and software. One interesting
finding was that PIs expressed a desire for time to learn how to
use existing resources. This latter finding suggests developing
orientation and training initiatives that could be conveniently
accessed by PIs.
The following section presents a more detailed
analysis of the responses while the last two sections provide the
survey questionnaire and a complete compilation of the responses.
Principal Investigator Survey Results
Principal investigators (PIs) were asked to
compare their access capacity to that of colleagues at other
institutions. While many elected not to respond to this item,
disadvantages were noted with sole-sourced paralleled resources
outside OSU, ISDN videoconferencing, and bandwidth. In response
to a question about being disadvantaged with respect to securing
research funding, lack of university funds for equipment, lack of
access to fiber optic links, and limited availability of local
programming expertise were noted as impediments. In terms of PI
expectations of their colleges and departments, most respondents
cited the need for increased spending on up-to-date equipment and
on technical support. Proper IT equipping of research labs was
also mentioned.
When asked what the university could do to
improve research-computing infrastructure, PIs suggested a variety
of enhancements, including greater return of indirect cost
recovery to departments and eliminating taxing research grants to
support infrastructure. Access to technical support was also
mentioned by several PIs, whereas increased bandwidth, an increase
in licensing agreements, and access to fiber optic links were each
recommended by at least one PI. Further, one respondent noted the
need for coordinated planning to build infrastructure across
colleges and departments and coordinated planning for technical
support personnel.
In response to an inquiry about “waves of the
future” with regard to research computing, real-time processing
with mixed modality input (e.g., video, speech), wireless links,
fiber optics, virtual reality and 3-D visualization, mining of
data repositories, GIS databases, and IP telephony were mentioned.
Questions about how the university was positioned for the future
listed the most and most varied responses. Whereas several PIs
expressed confidence about the universityˆs current position, it
was noted that we are behind in fiscal support, distance
education, the systematic upgrading of hardware and software,
classroom connectivity, and leveraging the expertise (e.g., CIS,
OSC, ACCAD, ITEC-Ohio) that we do have.
One questions asked how much sponsored
research money supported the PIs computing infrastructure, and
another asked how much the university cost-shared research-related
computing equipment. The findings indicate that the great
majority of funding is coming from grants as opposed to the
university, with a few notable exceptions.
When asked about other areas of interest, 9
PIs expressed a desire for time to learn how to use resources that
are currently available. Six expressed an interest in tera-scale
storage, and several PIs were interested in statistical analysis
software and consultation services.
While the most salient issue was clearly
university and college funding and support, the survey suggested
several directions to improve research information technology
support at OSU. These directions include improved training, better
coordination with the Ohio Supercomputer Center, development of
data mining capabilities and expertise, emphasis on virtual
reality and visualization, more access to qualified support
personnel, more access to information about technology, and more
available software, especially for data analysis.
RCAC Principal Investigator Information technology Survey
Survey Tool
- Please describe any examples of occasions in the past three years when your colleagues at other institutions had effective access to computing or communications technology that you did not.
- Have you been disadvantaged within the past five years in getting research funding because you did not have the computing technology resources (equipment, infrastructure, training, staff support) to be competitive? If yes, can you give an example? Does this disadvantage still exist?
- What can your College or Department do to improve the research-computing infrastructure to enable you to get more grants?
- What can the university do to improve the research-computing infrastructure to enable you to get more grants?
- What, in your view, are the waves of the future with regard to research computing and communication technologies?
- Is OSU well positioned for the future? Please explain.
- Approximately how much (in dollars) of your sponsored research money has underwritten your research computing infrastructure and research software in the past three years?
- Approximately how much (in dollars) of research related computer equipment has the university cost-shared on your behalf in the past three years?
- How important are the following you your research? The weighting scale is 1 to 10, with 1 being most important.
 Parallel processing
 Wireless networks
Virtual reality/Visualization
Data mining
Terascale computing
Other (please specify)
RCAC Principal Investigator information Technology Survey
Compilation of Responses
1. Please describe any examples of occasions
in the past three years when your colleagues at other institutions had
effective access to computing or communications technology that you did
not.
We simply could not finish or publish research on paralled computer graphics due to the lack of sole-source paralleled resources. OSC’s batch mode of operation is useless to us. We have since built our own and suffered all of the headaches associated with this.
Virtually every one of my colleagues across the country, at large & small institutions, has better access. First, they are supplied on a regular basis with up-to-date machines, and second, there are actually funds in the University budget to purchase peripherals for individual faculty.
Video conferencing w/ ISDN line requirements was not possible but would have been useful.
None in my case due to strong department resources, OSC & link to PACI & NERSC.
None. Maybe you should identify any novel technologies that are available at other institutions that can benefit the researchers here at OSU.
Computers behind & software behind
Access to desktop computing for classes with appropriate software
(eg. SAS, SPSS that can be scheduled for regular classes. Need about 30-45 stations)
OSU 26 K Band, PSU 46 K Band
2. Have you been disadvantaged within the
past five years in getting research funding because you did not have the
computing technology resources (equipment, infrastructure, training,
staff support) to be competitive? If yes, can you give an example?
Does this disadvantage still exist?
Yes, as a Research Center with a Systems Manager or even a system, we have been at a disadvantage, some individuals have survived on their own.
Perhaps. My NSF program says they will not purchase computers, it is the University’s responsibility. OSU says the reverse.
No, I have not. I try to acquire what I need. My needs are not very sophisticated though. I use pretty basic stuff and that is available everywhere.
Yes, I was not qualified and had difficulty accessing GMIC system.
Not per se, but I have spent tens of thousands of dollars building and funding my own systems with minimal departmental support.
Gaining access to high-speed fiber links on which to perform computer systems research.
Of course - research money must be used to buy all my computer equipment, including students and software.
No. Lack of department operating money and Univ/State matching funds is a much bigger problem.
All my examples have to do with with being able to do prototype work before writing proposals. We tend to support faculty after they get funding.
I was in need of programming expertise and had to hire my own programmer.
Well yes. No time or personnel to provide stellar web pages and other marketing.
3. What can your College or Department do to
improve the research-computing infrastructure to enable you to get more
grants?
As a Research Center we have little to no access to College Departmental assets.
Provide me w/ up-to-date equipment & technical support.
At this point we are taking money cuts and could use more storage space & CPU speed.
Develop a regular schedule OT upgrading.
Strong internal department network & service; head of department system or head of college system& chair of OSC hardware committed.
Support the 2nd Division of the College of Medicine & Public Health, both in terms of equipment and personnel.
Make them easily accessible to everyone. Maybe speed up the access time for faculty.
The most important thing I need is support for clusters of PCs in my research lab (Mech. Eng). This measure means efficient networking, set-up at servers (file and computing), reliable back up procedures, etc. By “support,” I also include advice on what to buy (hardware and software) as well as assistance with setup. Right now, my grad students and I struggle along without real knowledge.
Provide high capability computers/access-to-computers with real-time analysis.
An excessive amount of faculty time is wasted due to little or poor system administration and management of resource intensive research labs.
Spend money on it rather than just passing on costs to the department or PI. There is zero planning and commitment of SBS on this.
In our college technology infrastructure is less of a barrier than other support functions. Need more easily accessible information as to what is available and what is known about it, i.e. FAQs.
Provide an infrastructure that supports technology, including the provision of high end computers for faculty and technical support .
Work with OIT, OSC, and ITEC-Ohio to tear down barriers to access fiber links. We also need more research lab space as we hire new faculty.
I rely on OSC.
Remove 2% computer “tax” on grant direct costs.
Support for research centers that provide locations for especially expensive or knowledge intensive equipment/software and data.
Better manpower support: someone who knows how to troubleshoot/ solve s/w configuration problems.
Better back up systems for our system in the local network.
Prepare budgets and take care of all other red tape required by OSURF.
Very poor access to programming skills in
Dept. and College. Access to computers/e-mail/communications not a
problem.
Maintain an effective
computing infrastructure.
I process all
my data through the computer center at OARDC and they are very helpful
and supportive.
Improved access to
microcomputers for graduate research assistants.
4. What can the university do to improve the
research-computing infrastructure to enable you to get more grants?
The mainframe was not so bad. The move towards PC Windows network does not support central statistical expertise.
Need to acknowledge that this is an important issue and put funds toward solving it.
Provide me with up-to-date support.
Continuing to pursue and increase licensing agreements that will allow access to software.
Surely you joke. Get OSURF into 21st century.
Support the 2nd Division of the College of Medicine & Public Health both in terms of equipment and personnel.
Make them easily accessible to everyone. Maybe speed up the access time for faculty.
Provide access to super-computer.
Higher network bandwidth to the outside and dedicated high-speed bandwidth to OSC.
Greater return of indirect cost recovery to departments which can then be allocated to equipment needs and support staff.
Have a plan to build infrastructure across college or departments. There is zero planning for technical support personnel.
I think good grants people can do it with what we have available. Novices are out of luck.
I believe if I had a researcher or post-doc I could generate about 100K or more per year in grants. Our big limitation is grants. We are very fundable.
Make the access and support of technology more seamless (and readily available)
Provide the capability to use state-of-the-art levels of fiber for selected research in systems and networking. Provide additional research lab space (or other space so we can create new research labs in existing space)
The university focus is on providing internet access for students-largely music and games, ok?
Prevent departments from taxing research grants to support computing infrastructure.
Access to staff who can provide technical support when/as needed for special problems: code development/ integration.
Research computing infrastructure is not a constraint in getting grants (for me).
5. What, in your view, are the waves of the
future with regard to research computing and communication technologies?
I think we will go back towards mainframes, but with GUI interfaces.
Right now I only want a good machine that is reliable.
bIntroduce parallel/object-oriented, visualization, data-mining concepts when students walk in the door.
Very high power (speed/memory) desk-top computers
Real-time conferencing/interactions free or at low cost.
IP Telephones.
I can see study sections being able to “interview” grantees to get questions answered in real time.
Within education: 1) virtual reality and 3-D visualization, etc., 2) assessment of student learning via technology, 3) solutions to issues related to digital divide, 4) evaluation models appropriate to distance courses.
We seem a day late and a dollar short on our technology infrastructures and support.
Optical fiber, wireless, and sensor-based networks. Computation involving large data repositories, distributed across geographically distant areas.
High speed wireless links for rapid feedback during research “experiments.”
Telemetry/ datalink with enough bandwidth to support real-time weather updates of traffic info to aircraft in flight.
GIS, LandSAT.
Improve microwave/phone interactions; communication between Wooster and Columbus campuses.
Programming, bioinformatics, computational biology.
It is very difficult to predict. Thus, an effective assistance program must be maintained.
1) Internet-based dissemination of results, 2) Increase importance of Geographic Information Systems-bases data.
Real-time processing with mixed modality input - video, speech, gesture.
6. Is OSU well positioned for the future? Please explain.
I think we’ve gone too far towards PC’s and lost large-scale computing expertise.
No, right now I only want a good machine that is reliable.
OSU is a teaching institution. All students need to understand and be able to use computers-not all students need powerful computers.
OSC is well positioned. Get OSU out of my way.
I would think so. They are not the leaders but they are good followers. So right now they have what everyone else seems to have.
Not particularly-the main problem is the staff (faculty and A& P) all are overworked-hard to find time to get together and when together to get beyond the day to day to conceptualize a vision. TELR IS NOT A HELP HERE!
Regarding IP telephones, not that I can see.
No-go visit the University of Illinois and you will know what I mean.
Telemed. & distance access necks upgrading.
Our team is positioned to be the top in the country (on sustainable/ organic agriculture)!
No, I’m frustrated by the fact that the computer upgrades are put off until the point of being so outdated that work is stymied. In contrast, colleagues at other universities are continually offered on an annual basis hardware and software upgrades & are given both a desktop & laptop with the goal of productivity.
No. We are behind in distance education. We are not a player in the game and we need to get into the game in a much larger way.
In some ways we are. We have faculty research expertise in these areas, and have ITEC-Ohio next door. But to the extent that there are barriers in giving our researchers access to the technology required to perform state of the art research (and there appear to be such barriers), we cannot reap the true benefits of this expertise.
No way-we spend nothing on research computing and faculty computing resources.
No. Not enough research support. Too much overhead, and soon we’ll be taxed on our current operations!
No. We’re all stretched much too thin trying to “do more with less.” Time is a much more critical factor than equipment.
Don’t have any internet connections or power strips for student laptops in classrooms. Even instructor hookups limited. No interconnect for live interactions with students: on-line Q & A tabulations.
No-more into programming, bioinformatics, computational biology; Fast!
Within our own department we continue to update computers and networks as funds become available for students and general areas. Regarding individuals, many of us purchased laptops from outside funding for our needs.
I feel our department is well situated. We have a technician who helps all of us with computing problems and over at OARDC have good support in the statistics lab.
Much better now than a decade ago. Still, OSU is far from the leading edge in computing resources, provision of site licenses for analytical software, etc.
We have great resources in CIS, OSC, Medical School, ACCAD, but lack the people resources to leverage these fully. As well as the funding (endowments, etc.) to sustain these.
7. Approximately how much (in dollars) of
your sponsored research money has underwritten your research computing
infrastructure and research software in the past three years?
$10,000.00
As a center as a whole, my estimate would be $50K yr.
$0
$3,000.00
$2,500.00
We write all our software, I buy all platforms on post-doc/ students desks in my group. I’d say $75,000 (underestimated).
Approximately $200,000.00
All of it. University has provided the access to internet and that is a big help. But, then that is the least they could do.
$4,000.00
$5,000.00
Well in excess of $1 million.
$60K
Less than 1%.
$0.
$20,000
$186,799
$100,000
$30,000 plus $10,000 for hardware and software.
Almost none unless you count data collection costs. In that case about $30,000 to $35,000
11 out of 25 K in just one case.
Very little direct money; quite a bit indirect (i.e. students programming to develop analysis systems)
$25,000-$30,000 (new server and some small stuff). Plus I purchased 2 new computers (PC/Mac) for my office with money for mentoring an NSF post doc!
Very little
None-most federal grants do not permit this
$75,000-100,000
Approximately $7000 annually
100%
$5000
Approximately $500,000
Less than 10%
8. Approximately how much (in dollars) of
research related computer equipment has the university cost-shared on
your behalf in the past three years?
$0
In reality, none except if you count start-up funds for individual researchers.
$0
All of it.
$2,500.00
Zip.
$0
I am tempted to say none. But that won’t be accurate. As I said, they have provided the access.
$0
$0
None.
$0
99%.
$0
$423,046
$0
$<1,000
$5,000
None.
ZIP- all but the desktop machine in my office came from funded research, NOT this university!
None
0
I bought the present computers with start-up money from the college
None. I purchase through grants (about $7000/year) out of about $150/year in grant support for my program.
O%
$5000 was used to buy new computers out of soft money
None
About $200,000
$3000
9. How important are the following you your
research? The weighting scale is 1 to 10, with 1 being most important.
Parallel processing
Response: 1 2 3 4 5 6 7 8 9 10
No. of responses: 2 3 2 0 3 1 0 2 0 2
Wireless networks
Response: 1 2 3 4 5 6 7 8 9 10
No. of responses: 1 1 1 3 4 0 3 0 1 3
Virtual reality/Visualization
>Response: 1 2 3 4 5 6 7 8 9 10
No. of responses: 3 2 1 3 2 1 0 0 0 5
Data mining
Response: 1 2 3 4 5 6 7 8 9 10
No. of responses: 3 2 3 1 6 0 0 1 0 1
Terascale computing
Response: 1 2 3 4 5 6 7 8 9 10
No. of responses: 0 0 4 0 1 0 0 2 1 4
Responses of particular note:
-One person recorded “1” for all five areas
-Two people said, “All are important areas to me”
-Two people recorded “10” for all five areas
-One person circled “Parallel processing” and “Virtual reality/Visualization
-One person checked off all five areas, except recorded “10” for Wireless networks
Other areas of interest:
Central Computing facility with consulting for statistics.
Object-oriented, high-performance (1); Virtual reality/Visualization & Data mining are the same to me.
Statistical Software (1).
Support to have the time to learn how to use some of what’s available(9).
My needs are normally just analysis of a batch of data taken at the end of each year on each experiment and then a summary analysis for publication.
Need real-time control and analysis systems
GIS
Statistical analysis (1)
Software design (1), Image processing (1), Interfacing (1)
Access and collection of large-scale data sets (1), Tera storage (6).
Statistical software (1)
Support to have the time to learn how to use some of what's available (9).
Other Notes:
One person did not fill out the survey, but wrote:
"PIs get basically no support for the IDC paid. In addition, my
contracts charge a computing fee for my graduate students, but the fees
pay only for undergraduate classroom support, not the 999 enrollment.
My department pays approximately $3.5 M annually to IDC; the best
computing solution is to return a portion to the units to hire IT and
computer site staff for research support."
Return to PlanIT home
|
