|
|
A Proposal
|
This web page is http://www.bernstein-plus-sons.com/.dowling/academic_computing_proposal.html
Copyright © 2002 Herbert J. Bernstein. All rights reserved.
We propose that, in order to enhance computational resources for instruction and research at Dowling College while preserving the integrity of administrative computing systems, Dowling College return to an approach it once followed and separate academic from administrative computing. In making this recommendation we recognize and applaud the valuable contributions made to computational resources by Dr. Black and his subordinates, but note that, just as the school is reflecting upon the future of its administrative information processing infrastructure in Project Delta, reflection upon the future of its academic information processing infrastructure is necessary and appropriate. The difference is that recent events mandate that any adjustment on the academic side be done much more rapidly than is being contemplated on the administrative side. By decoupling future efforts in academic and administrative computing, the fine work done to date can bear fruit in both areas despite their different needs.
One should not infer from Dr. Black's process reengineering effort that there is some serious defect in the current infrastructure. On the contrary, Dowling College has a sound and well-managed network, and a knowledgeable and supportive staff. The administrative systems function and are an essential part of the operation of Dowling College. Thus it is reasonable and prudent for Dr. Black to be planning a slow and careful review of these systems over the next two years to position Dowling College properly for whatever changes and improvements in administrative systems may be required.
However, on the academic side, rapid review and rapid change are essential. We need to convert a bulge in applications into a bulge in enrollment. If we are to make that work, we must have classrooms in which to teach this influx of students, and those classrooms must have the appropriate infrastructure. In many cases the necessary infrastructure includes computers and network access for instructors and, in some cases, for the students as well. The simple internal network access now provided with one-way internet access was a good start for academic computing, but must give way to a full embrace of the internet for the 21st Century. This must be done and done rapidly, but must also be done without compromise to the integrity of the administrative systems. Only by completely decoupling these efforts can we achieve both the changes the academic side needs and the security the administrative side needs.
Dowling College is in a time of unique challenges and opportunities. Applications are up dramatically at Dowling and at other schools in the area. In order to convert this bulge in applications into a significant long-term increase in enrollment and retention, the College must be able to respond to the needs of its students and faculty on very short time scales. Many of those needs involve use of computers and networks. Dowling College has made major investments in computational infrastructure in support of administrative activities, instruction and research. The Dowling administration is to be complimented on its vision and foresight in providing these resources. Now, if it is to absorb many new students, it must make rapid improvements to the infrastructure in support of instruction and research. The time scales that are appropriate to planning for administrative systems are far too long to allow us to take advantage of the bulge in applications. In this document we propose a division of labor between academic and administrative computing to build on the past investment in ways that should help Dowling to respond in a rapid, cost-effective manner on the academic side to remain competitive with other schools in the area while preserving the integrity of essential administrative systems.
In the realm of computing and networks, times and technology are changing rapidly. Most importantly, the pace of change is increasing. Computers are increasingly a disposable commodity item. Unlike administrative networks, academic networks require ever-increasing capacity and need to be made available throughout the campus. Hardware costs continue to decline, but increased staffing is needed for computer-related activities. Security issues that were minor nuisances a few years ago now threaten computer access throughout the internet. If Dowling is to compete effectively for students and to provide faculty with the environment they need to teach effectively, the school must become increasingly adaptable and agile in its technological activities related to instruction and research. At the same time it must retain its robust and stable administrative systems. The need for rapid change and openness on the academic side is in direct conflict with the need for security and stability on the administrative side.
The time has come for Dowling to return to an approach it once followed -- the strict and formal separation of academic computing from administrative computing. Only with a clean separation can the needs of both areas properly be addressed.
Technology has become a major investment for most colleges. Prudence and conservation of resources has led to careful management of technological resources at the highest administrative levels. Financial systems, voice and data communications systems, student and faculty computational support have all effectively been combined. This has helped in arranging resources to wire the campuses and to establish computer laboratories.
The unfortunate events of 11 September 2001 have made many students apply to Colleges closer to home. If we accept more students than we can handle well, we will damage Dowling's reputation and have trouble with retention. Many components of Dowling's infrastructure will have to grow rapidly to allow us to accept students and give them an academic experience that will make them wish to stay at Dowling. We cannot afford the luxury of months or years of lead-time in arranging such facilities. When newly enrolled students walk into class in less than six months, everything needed to teach those students must be ready.
In the 1990's it could fairly be claimed that the computing needs of administrators and of faculty had much in common. Wintel style PC's dominated in both areas. UTP wired Ethernet networking dominated in both areas. Learning of Microsoft Office suites was important in both areas. Individual computers, network infrastructure and software licenses were major cost components in budgets that had to be managed with great care. It was expected to make use of the same PC for several years. It was reasonable to use planning processes for network drops that might take months or years. Most uses of computers were internal to the school with modest demands for external traffic. It was reasonable to think in terms of a single, centrally managed school web site.
In the 21st century, while many of these factors remain valid for administrative uses, for academic uses the parameters have changed. Unix has become an important factor in many areas of instruction and research. Mac OS has successfully fought back against the push by MS windows to extinguish it, and no school can do a credible job in teaching or using multi-media tools without a Mac OS presence. Costs for viable student computers have dropped to well below $1000 per student machine for notebook and desktop computers. Costs for networks are also dropping rapidly, especially with the introduction of wireless Ethernet network infrastructure. While a given student might make effective use of the same computer for 4 years, the rapid pace of technological change makes it desirable for instructional computer hardware to be upgraded every year, and for software to be upgraded at least once per semester.
Students need access to the internet to prepare for classes and to communicate by email. Students need computers and printers to prepare papers. Students need to learn to use the worldwide web to present their work. Students in the natural sciences and social sciences need computers to run software peculiar to their disciplines. Students in more and more areas need computers as essential tools of their education. Faculty need computers for similar reasons and to perform various administrative functions.
Changes in the realm of instruction and research must be addressed on appropriate time scales. The administrative side of Dowling has, in Project Delta, committed itself to a thorough examination of its administrative information-processing infrastructure. This is a laudable effort, timed, as is appropriate for such administrative efforts, to extend over 2 years. However, at exactly the same time, if Dowling is to adapt to the likely changes in its enrollment and demographics, the academic information processing infrastructure must also be reengineered, not once, but several times over: first between now and Fall 2002 to be ready for the influx of new students from the bulge in applications, and then again for each semester thereafter to help create an environment that will help to retain these students and attract new ones.
For students, the implications are clear: students have an ever-increasing need for access to computers and networks, and the need for rapid change of student systems and the need for security of administrative systems argue strongly for a clean administrative and technical separation between student-related academic computing and administrative computing. It is administratively unsound to couple budgets and technically unsound to couple networks for such different activities. Students do have a need for some access to administrative systems, but security concerns mandate that most such access either be mediated through a Dowling office, or be handled indirectly over the web, rather than by providing students with direct access to administrative systems. Where direct access is to be granted, legal requirements for protection of student records and common sense prudence about security of financial systems argue for carefully managed limited access without assumptions of trust of the connection or of the user.
For faculty there is much in common with the situation for students, but the administrative needs of faculty to access student records and financial systems requires a certain degree of interaction with administrative systems. Fortunately, increasing security threats and the technology developed to deal with them have provided solutions that have diminished or eliminated the need for faculty networks to be integrated with administrative networks. This can be seen in the web interface to on-line student records which is accessible to faculty over the worldwide web in the same way that portions of the system are available to students. If the same methodology were extended to allow faculty access to certain appropriate functions within the financial systems via the internet, then any need for coupling between faculty network access and administrative network access would cease.
Since both students and faculty have needs for network and computer access that conflict in timescale and security requirements with the timescales and needs of administrative systems, it is our recommendation to Dowling College that a clean separation be made between academic and administrative computing, with academic computing being in a line reporting to the Provost, with its own network, budget and staff.
To achieve this separation some persons now handling academic computing matters under Dr. Black would move under a newly designated Associate Provost for Academic Computing (APAC). It is essential that this person be a respected academic with teaching experience, who would remain actively involved in teaching. The cooperation of the Union should be sought in making such arrangements. One would hope such cooperation would be forthcoming, since the requirement is intended to ensure that the APAC is sensitive to the needs of faculty working in a classroom setting.
Existing equipment and network infrastructure would need to be partitioned between academic and administrative computing. Fortunately, recent changes in technology have made it relatively inexpensive to provide duplicate and/or alternative network infrastructure. At worst a new LightPath cable drop at Oakdale and Brookhaven would be needed for external academic computing access, and several wireless internet access points might be needed to help bridge gaps in network infrastructure where critical legs have to remain with the administrative side and new wired drops could not be installed in time.
A fortunate side-effect of making more use of wireless internet is that it would become practical to equip many more classrooms with network access, thereby providing much more freedom in room assignments during the difficult period expected in Fall 2002. By adding wireless access points and using portable projectors, we can make many classrooms computer-ready on very short notice. For a class that only needs an internet-connected computer and projector for the instructor, the cost can be less than $10,000 per setup. Simpler setups using projection panels on overhead projectors can be done for less than $6,000 per setup. For a class that needs computers for the students, the cost can be less than $1000 per internet-connected student machine. Such setups do not require major changes to the physical plant and can be done very quickly. Dowling College already has examples of such equipment. What is needed is an organization and budget to give higher priority to such efforts.
Indeed, if current cost trends continue, it may be most cost-effective to hand every incoming freshman a laptop with a wireless card to keep and provide wireless access points in dorms and classrooms, rather than to invest in new desktop computers that will be obsolete and have to be replaced the next year. Non-freshman who have similar needs could be offered similar machines and/or wireless cards at attractive prices. Such steps could go a long way towards making the educational experience at Dowling highly competitive, and help in retention.