Understanding The Changing Forces In Information Technology |-------------------------------------| | Paper presented at CAUSE92 | | December 1-4, 1992, Dallas, Texas | |-------------------------------------| UNDERSTANDING THE CHANGING FORCES IN INFORMATION TECHNOLOGY Carole Cotton CCA Consulting Inc. Wellesley Hills MA 02181 ABSTRACT Many schools are decentralizing management authority. There is a clear trend away from institutional decision making and toward departmental authority. Implicit in this new management model is the need for distributed access to information. With distributed access comes the need for new computing models. How fast are the changes occurring? Will major applications be downsized? And, what about "open" systems and client-server computing... What impact are they having on IT in higher education? This paper will examine several of the changing IT forces currently affecting higher education. The data is derived from our 1991 and 1992 surveys of higher education. INTRODUCTION Our 1991 and 1992 Studies entitled Understanding Information Systems in Higher Education reported the results of extensive surveys into many aspects of IT on the campus. While it is not possible, in this short paper, to summarize two 200-page reports, we will discuss several key aspects of interest to both educators and technologists. The data is taken from the actual survey responses and is augmented by comments obtained by interviewing many notable trendsetters in higher education. We begin by looking at changes to the computing paradigm. COMPUTING PARADIGM It seems that everyone is talking about changes to the computing model... Open systems and client-server architectures: what do these terms mean to institutions and how fast will the changes occur? Let's look at how our respondents answered some of these key questions. [FIGURE MISSING FROM TEXT FILE] According to our respondents, we see that "open" really means freedom of movement -- across platforms and across networks. Transportable software, interoperability and open network protocols were the most frequently mentioned responses. But, perhaps the most telling insight from this data is that no single answer was mentioned by more than 62% of the respondents. Does this mean that over one-third of the respondents do not believe that transportable software and interoperability are determinants of "open systems?" The format of this question was not restrictive. Respondents could have selected all options... but they did not. Why? Instinctively we all know the answer. The overall IT industry has not yet -- and may never -- achieve consensus on this topic. Leaving the users with only one choice: to define "open systems" within their own unique, contextual framework. UNIX: MYTH VERSUS REALITY [FIGURE MISSING FROM TEXT FILE] Assuming that our respondent base reasonably represents the higher education universe, then it appears that "UNIX Power" has grown more slowly than expected. Forty-eight percent of the respondents reported that they had no systems whose operating environment was UNIX... This means that there are likely to be as many institutions that do not have UNIX systems as there are those who do! Further, for those that do have UNIX on campus, it often fills a specialized role. Twelve percent reported UNIX on servers only and another eighteen percent only on workstations. Further, only eight percent currently see UNIX as a (as opposed to the) preferred operating system. WHO USES UNIX? One would expect to see marked differences in the use of UNIX across the Carnegie classes, and our data supports that thesis. Universities have a much higher penetration of UNIX than others. With 100% UNIX penetration, Universities are clearly leading the march toward UNIX. Four-Year schools are next in line, with almost two-thirds reporting a UNIX presence, and are followed by Specialized institutions. Two-Year schools -- with only 30% UNIX penetration -- appear to have the lowest adoption rate. [FIGURE MISSING FROM TEXT FILE] While UNIX does not have broad-based penetration, it does appear to have won the confidence of those who have embraced it. Eighteen percent of respondents said they had a installed a major (mainframe or minicomputer) Administrative or Library system whose operating environment was UNIX. That's a major commitment. The reasons for their decision shed still more light on the role being played by UNIX. In a related question, we asked why they purchased a UNIX system. Sixty-five percent replied: the application ran on UNIX; and over forty percent mentioned price/performance of the total solution. In contrast, only eighteen percent mentioned an institution-wide commitment to UNIX. It does appear that UNIX-based Administrative and Library system decisions are driven more by local decision makers "betting" on the total solution -- and less by institutions "betting" on UNIX. Examining the flip-side of this coin, we asked those who had not installed a major UNIX-based system: "why not?" Fully half mentioned incompatibility with existing systems, and nearly an equal number felt they lacked in-house UNIX expertise. Further, the data does not suggest that schools are avoiding UNIX because they lack faith in the operating system. Common trade press complaints like security issues and inadequate development tools were mentioned infrequently. Rather, it is the need to fit into the current legacy systems and the lack of resources that seem to represent the foremost obstacles. It does appear that if schools believed that a variety of quality applications existed, and that they were compatible with existing systems, the rate of UNIX adoption might be very different. NO CONSENSUS ON CLIENT-SERVER [FIGURE MISSING FROM TEXT FILE] We asked: what does client-server mean to your institution?. Summarizing their answers, we see that a majority of respondents indicated that client-server computing means that two systems cooperate -- in some manner -- to execute a task. In keeping with the pattern of answers for "open" systems, we see that forty percent did not agree with any of the proposed models. Client-server definitions, like those for "open" systems, share a relatively low level of consensus. In reality, the only area of agreement is that neither application "downsizing" nor the networked terminal approach fulfills the definition of client-server computing. And, while there are many differing points of view with regard to "the" best approach to client-server computing, it still enjoys considerable popularity. This emerging desktop-centric and network-connected computing model is much closer to the faculty work model than is a central computing operation, and it is being embraced with significant enthusiasm. WHO PLANS TO IMPLEMENT CLIENT-SERVER? [FIGURE MISSING FROM TEXT FILE] How many institutions are planning to implement client-server applications? MOST! Overall sixty percent said they were planning for the implementation of client-server applications. There are, however, significant differences across the total population. Ninety-six percent of the Universities; 65% of 4-Year; 51% of 2-Year and only 43% of Specialized institutions indicated plans for implementing client-server applications. The pattern by size of institution is just as strong: eighty-six percent of the largest schools have plans for client-server applications, in contrast to the smallest schools, where only 41% have similar plans. Clearly, most institutions are looking to move to client-server computing. The next question is why. We asked respondents to describe their primary motivation(s) for moving to client-server computing and what impediments might stand in their way. As to motivations, we found that "increased functionality" and "price/performance" were mentioned most frequently. In contrast, the list of the industry-wide factors such as "hardware scalability" and "industry trends," fell to the bottom of the ranking. The decision to move to client-server computing -- like that for UNIX -- is based on micro-level issues of need fulfillment and not on macro-level issues like molding the campus IT infrastructure to fit the changing computing paradigm. Just as the motivations mentioned for moving to client-server mirror those for UNIX implementations, so do the impediments. "Money" and "staff/user skill sets" were mentioned by over three-quarters of the respondents. "Network infrastructure" issues were also mentioned as schools question whether their existing networks can support the increased network load brought on by client-server computing. Of note, however, is that none of the aforementioned issues is a specific indictment of client-server computing, but instead reflects the problem of moving a large existing base of hardware, software, and skill sets to a new and different paradigm. No one expects the move to client-server computing to be without some level of pain, as most see the need for significant hardware realignment. Close to half of all schools said they expect the introduction of client-server computing to require both re-configuring existing equipment and retiring incompatible equipment. WHAT'S CONNECTED? [FIGURE MISSING FROM TEXT FILE] Has the anticipation of new computing models had an effect on overall campus connectivity? The answer appears to be yes. Starting at the top, there have been large increases in connectivity -- for all space categories -- over the past year. Faculty/Staff offices and Libraries enjoyed the highest rate of connection in 1991. These two areas maintain their top place in 1992, but with considerably more connections having been installed over the past year. Faculty/Staff offices have increased to 52% and Libraries to 51%... And for the first time, we find some campus space categories with more connected to the network than not. Of special note amongst this upbeat picture of network connectivity, is that of Student dorms. In spite of a large gain, they still trail all other types of space in access to the campus network, and this figure represents only those students living on campus. Missing entirely are the vast numbers who live off-campus. Indeed, many schools are trying to address this issue, but until everyone, including those students and faculty who live and work off-campus have equal access to the network, the dream of universal connectivity will remain unfulfilled! CONNECTIVITY BY CARNEGIE CLASS [FIGURE MISSING FROM TEXT FILE] Does connectivity vary by Carnegie class? Universities offer the widest array of network services. Do they also have the highest rates of connectivity for all space categories? The short answer is no. Universities do lead in network connections to Libraries and for "other" types of space, but fall behind in all other categories. The differences are not large, but they are only in second place for Student Dorms and Faculty/Staff offices. And are actually in last place in wiring Classrooms to the network backbone. In contrast, 2-Year schools lead all others in classroom connectivity. CENTRALIZATION... MORE OR LESS? [FIGURE MISSING FROM TEXT FILE] There is much discussion about distributing decision making in schools -- and with it -- the computing infrastructure to enable this change. As a result, one might expect to see a movement away from centralization of IT staff organizations toward a more decentralized model. We tested for the current and planned level of centralization. Respondents were given a choice of answering "now," "plan" or both. (Please note that a drop- off between "now" and "plan" does not imply declining centralization.) The results indicate that schools are planning for more -- and not less -- centralization of computing operations. Of the respondents who indicated that they were planning for more centralization, approximately fifty percent said they planned to include Academic computing and Libraries in their central organizations. Does this mean that increased decentralization of computing power will result in even more centralization of IT staffs as schools realize that they need central functions to ensure minimum standards for data security, network standards, data architectures, as well as their overall IT architectures? [FIGURE MISSING FROM TEXT FILE] Switching now to a view which describes only the current level of centralization, we see that organizational reporting does vary by Carnegie class. And, some types of organizations vary more than others. Over eighty percent of all schools indicated that Administration reported centrally, and the differences vary only slightly by school type. Academic computing is the next most likely function to report to a central organization. Overall, sixty-five percent report centrally, but here we see variances across the Carnegie classes. Universities and 4- Year schools are above the mean response, and 2-Year and Specialized institutions are below it. It appears that one in four Libraries report centrally, but the range varies from a high of just under forty percent for Universities to a low of 20%, for Specialized institutions. Currently, about one in four Computer Science departments are likely to report centrally, and 2- and 4-Year schools lead the way, while Universities and Specialized schools anchor the low end of the curve. Finally, Engineering is the least likely -- for all Carnegie classes -- to report centrally. ... So why aren't more computing operations combined? Politics! In a related question, we asked respondents to define the major impediments to combining IT organizations. Money (or lack thereof) was mentioned most frequently, but when the responses for "institutions philosophy" and "culture" were combined, we discovered that over seventy percent of the respondents said the impediments to combining IT organizations were mired in organizational morass -- translate politics! Further, all types of schools mentioned the "political" issues. The span of responses ranges from a high of over 90% for Universities to a low of sixty percent for 2-Year schools. Personnel skill sets and all other impediments like data security, network infrastructure and remote campuses, etc. were mentioned by far fewer respondents. The bottom line here is politics, the rest is commentary! STAFFING... UP, DOWN OR FLAT? [FIGURE MISSING FROM TEXT FILE] The general scenario at most schools calls for falling -- or at best -- flat budgets. What impact is this having on staffing for central IT organizations? Our data suggests that overall staffing in both Administrative & Academic computing organizations is up slightly from 1991, but that there are wide variances by institution type and size. Looking first at the overall picture of the mean number of full-time equivalents (FTE's) for all schools, we see that the 1992 figure is eighteen, an increase of four FTE's over the 1991 figure of fourteen. However, the increases are not spread evenly across all schools. Clearly, Universities have experienced the highest level of staffing cuts. Mean University staffing levels dropped from eighty FTE's in 1991 to seventy-one in 1992. Specialized institutions have also dropped slightly from eleven in 1991 to ten in 1992. Counterbalancing these cuts, however, are the increases for 2- and 4-Year schools. Four-Year schools almost doubled their mean FTE headcount as they moved from nine in 1991 to seventeen in 1992. The 2-Year schools also moved upward from nine to ten. Further, as we looked below this surface level and compared staffing changes by size of staff (as opposed to Carnegie class or institution size), we learned that all Carnegie classes -- with larger staffs -- experienced staffing decreases. Specifically, there were staff cuts for all schools with staffing ranges of fifty or more FTE's. IMPACT ON LIBRARIES Is the library being affected by all these changes? The basic mission of the library is not changing, however user needs are changing and many of the strategies employed in its service delivery are being altered, some in a dramatic manner. More information is being stored and shared -- across the campus and around the world. Further, more people are accessing the same information... in different locations... at different times... and for different purposes. Since changes in strategy are usually accompanied by changes in structure, let's look at some of the most prominent structural changes. [FIGURE MISSING FROM TEXT FILE] With increased demands for on-line access to information, we would expect to see significant change in the level of library connectivity required to enable these changes... And we did. Currently, almost half of all schools provide network access to the library catalog, and this is almost twice the rate we saw in our 1991 Study. However, the picture varies considerably by Carnegie class. Universities lead the way, with ninety-five percent offering library services to their network users. The 4-Year schools reflect the industry average, and about one half of them offer this service. In contrast, the 2-Year and Specialized institutions are well below the industry average, with only one-third reporting network access to the Library catalog. CONNECTING LIBRARIES TO EACH OTHER [FIGURE MISSING FROM TEXT FILE] With half of all schools offering network access to their own library materials, the next likely step is resource sharing with other institutions. There is trend toward formal cooperation among libraries in order to provide better access to scholarly materials, and in some cases, to enable individual institutions to restrict the breadth of their own collections. One test for growth in this resource sharing thesis would be to find a significant increase, over the last year, in the number of institutions implementing the Z39.50 NISO standard. It provides for transparent exchange of data between systems that use different hardware, applications software and operating systems. And we did find marked increases across all Carnegie classes except for 2-Year schools. SUMMARY The general consensus is that the entire realm of information management is changing -- and at a very rapid rate. We see a coincidence of new management and new computing models, with information now being viewed as a prime enabler in meeting institutional goals. Further, technology needs are becoming more complex as more users -- in more disciplines -- use new forms of data like voice, text and images... And the rate of technological change continues its onward march. Yesterday's solutions -- or even today's -- will not meet tomorrow's needs... In this environment, the institutions that will benefit most from IT are those who are not just embracing change but, instead, are actively planning for it and driving it.