copyright notice
accesses since September 18, 2001

A SWOT Analysis of Computing as an Academic Discipline

Hal Berghel,
Prof. and Chair of Computer Science,
University of Nevada at Las Vegas

ABSTRACT: SWOT analyses are in widespread use, but haven't caught on in the academic computing and information science communities to monitor their health and well-being. In this paper we will deal with the Strengths, Weaknesses, Opportunities and Threats that face the academic CIS community in the United States. While a complete analysis is beyond the scope of this paper, I will present enough data to wet one's appetite for a more thorough-going investigation.

We will then turn to results of a recent survey of Computing and Information Technology Colleges in the U.S. conducted by the author will be summarized, and predictions will be made on forthcoming trends and paradigms in CIS.

STRENGTHS: The academic computing discipline in all of its manifestations is as far as I can determine without peer in terms of student interest and demand for graduates. According to the latest Taulbee Survey of U.S. computer science and engineering departments [1], production at the bachelors level has more than doubled in the past five years, while the number of declared majors has risen by 250% (see Figures 1 and 2).


Figure1: Computing BS Degress
(U.S.)1995-2000


Figure2: Declared CS/CE Undergraduate Majors
(U.S.) 1995-2000

There is no evidence on the horizon that these trends will reverse any time soon. One of the strengths of the computing discipline is the size of the interested student population.

WEAKNESSES: Ironically, one of the weaknesses of the computing discipline is the size of the interested student population as well. This unbridled increase in students comes at a time when PhD production has hit a ten-year low. Compare Figures 1 and 2 with the Taulbee data on PhD production in Figure 3.

Figure3: PhD Production in Computing
(U.S.) 1990-2000

Thus, while the demand for graduates continues to increase commensurate with the increased student interest and matriculation, the availability of new faculty to teach is dropping. In fact, U.S. PhD production in computing actually decreased this past year by 6%!

Another weakness of a very different sort deals with the current level of research support. While information technology continues to dominate the global economy, funding from the National Science Foundation for Computing and Information Science and Engineering for the current fiscal year is actually 1.6% lower than last year. What is wrong with this picture?

THREATS: For narrative reasons, I'm going to take the last two parameters of my analysis out of order. The reasons will become clear in the next section.

The thesis that I want to advance is that it is quite possible that the greatest threat to computing is the lack of understanding of computing and information technology as academic disciplines. To illustrate this point, I'll refer to a recent survey I conducted of the Deans of Colleges of Computing and Information Technology that are represented at the Computing Research Association's Workshop, held first at Snowbird in July, 2000, and then subsequently in Atlanta in March, 2001 and Berkeley in August, 2001. The survey results are illuminating.

Figure4: Formation of Colleges of C&IT

Figure 4 illustrates the trend toward freestanding colleges of computing. Notice that the as many colleges of computing were formed in 2001 as in the previous five years, and almost twice as many were formed in the past six years as in the three decades preceding. The question I sought to address in my survey was whether this was an ephemeral trend or a paradigm shift in the way the at computing and information technology programs would be organized and administered in modern universities.

If there is a general paradigm shift under way, it behooves us to look further into the rationale behind it and judge whether the objectives of the academic units that are re-defining themselves in this way are being realized. In a sentence, we seek to make a first pass at determining what our profession is trying to accomplish, and whether it is succeeding.

My survey was entirely Web-based. The questionnaire consisted of four parts:

  • general Information about the CIT unit,
  • questions about the prior status of the CIT unit before it became a separate college,
  • questions relating to the motivation for forming a separate unit (vs. remaining a department or program), and
  • questions relating to the current administration of the college.
    For those interested, the original survey instrument remains online at www.acm.org/top/coc/coc-ql.html, and the results of the survey are available online at http://www.acm.org/top/coc/IT-school-survey.html. I will only cover the highlights here, since the details are available online.

    One of the more interesting questions has to do with the original locus of the CIT colleges before their separation. The breakout was as follows (N=14): Engineering (11.76%), Sciences (23.53%), Business (5.88%), with 58.8% listing their origins as "other". These "other" origins are listed below exactly as they were entered in the questionnaire:

  • Library School
  • Interdisciplinary
  • Academic programs in Arts and Sciences
  • College of Applied Science and Technology
  • Arts & Sciences
  • Mathematics
  • Science and Liberal Studies
  • Graduate School
  • Applied Science and Technology

    This variation suggests to me one of the most likely problems underlying the appreciation of computing and information technology as an academic discipline - it is suffering from an identity crises. The very ubiquity of computing is at the same time its downfall, because anyone can claim ownership. When I remark at keynote talks that everyone who has ever successfully booted a Macintosh computer is by definition a CIT expert, the theme usually resonates with the audience. The breadth of the prior homes of these Colleges eschews any pretense of cohesion.

    So if computing can be anything, why will it fail to thrive anywhere? Some of the answers lie in the responses received on why the departments of computing decided to secede from their original homes. Figures 5 and 6 suggest some reasons.

    Figure5: Before thre separation, how would you compare your
    dapartment with other departments within the college in terms
    of credit-hour production

    Figure6: Before the separation, how would you compare
    your dapartment with other departments within the college in
    terms of the number of majors

    The point that emerges is that departments of computing tend to be the "cash cows" for their respective divisions. This is not to deny that there are philosophical reasons behind the separation of CIT programs from other colleges, but the over-productivity when compared to sister disciplines clearly plays a part.

    But what are these philosophical differences? Consider the response of the Deans regarding the external assessment of computing research.

    "Before the separation, how would you characterize the College's overall assessment of computing faculty research?" Responses (N=12): Poorly regarded, 50%; On a par with other departments within the College, 25%; Respected, 17 %; Well Respected, 8%."

    That 50% report "poorly regarded" suggests that we computing and information technology disciplines may still be suffering from a cultural disconnect from more traditional disciplines with which they have been historically associated. This phenomenon has of course been widely discussed in the literature [2][3][4].

    In addition, the Deans were asked to respond to five "hypothetical" motivations behind the separation of their computing programs from their previous colleges. Here are their responses.

    1. "starvation syndrome - unfavorable/inequitable resource allocation within existing university structure. Responses: of slight concern, 15%; of moderate concern, 23%; of critical concern, 62%.
    2. "overblown department phenomenon" - size dwarfed sister departments without commensurate increase of influence/resources. of slight concern, 38.5%; of moderate concern, 23%; of critical concern, 38.5%.
    3. "turfing" - more established departments refused to relinquish resources even when justified by economic reality. Responses: of slight concern, 23%; of moderate concern, 23%; of critical concern, 54%.
    4. "zero-summing problem" - needed reallocation of resources politically infeasible. Responses: of slight concern, 23%; of moderate concern, 23%; of critical concern, 54%.
    5. "mission-breadth" - unit had broader mission than peer departments than supervising unit could accommodate, and difficulty in migrating to interdisciplinary focus. Responses: of slight concern, 14%; of moderate concern, 0%; of critical concern, 85%.

    If I am correct, the threats to computing come largely from within their own organizational units - whether Colleges, Divisions, Universities or University Systems. The pattern seems to be that the larger institutional units are insensitive to the needs of the computing discipline, and unwilling to provide funding at levels commensurate with their contribution to the institution's mission.

    OPPORTUNITIES:

    I will leave you on a positive note. That the future of computing and information technology as a discipline is bright is beyond dispute at this point. What is conjectural is whether we as a society will treat the discipline in ways that will deliver maximum economic and social impact. Decreasing research spending, ignoring the needs of a student population that doubles every half-dozen years, failing to take appropriate measures to entice our best and brightest to enroll in PhD programs, and other such "evasive" actions will not get us where we need to be. As my survey suggests, some Colleges in the U.S. and Canada have determined that the only way to deal effectively with these problems is to form free-standing academic units.

    So what will help us achieve the desired effect of strengthening our global economies and building highly skilled and highly compensated workforces. The first answer must lie in introspection. What is it that we as a profession fail to do that accounts for our malaise? Some of the answers are suggested by Peter Denning in his observation that "it is an irony that the computing discipline, which gave birth to the IT profession, is not the driving force in the profession." [5] Whether our role is to be the custodian of the intellectual core of computing, or what I feel is more likely, to serve as the facilitator in the creation of new intellectual synergies, is difficult to determine. However, what is becoming quite clear is that we are not in enough control of our own fate. Within existing bureaucracies, we are infrastructurally-challenged from within and without. Within our own spheres of influence, we seem incapable of communicating the ubiquity and centrality of our discipline to modern intellectual life (hence the Macintosh example I alluded to earlier). Without our own spheres of influence, we seem incapable of effectuating adequate organizational change. In a sentence, policy makers want to derive the benefits of our labors without adequately nurturing our environment. Therein lies the rub.

    The beauty of the opportunity that I want to emphasize lies in its simplicity. Where other disciplines have to deal with continuous, and in many cases intractable, long-term problems such as lack of student interest, low economic rewards for professional effort, dwindling economic opportunities, etc. computing faces none of that. Our real problem is lack of understanding and misperception (even amongst ourselves!), and those are far easier to deal with than economic realities.

    Our opportunity is to communicate loudly and in harmony the obvious fact that our discipline is the intellectual glue of business, industry, government and academe. The fact that we face the challenges I have discussed lies in the fact that we fail to present a unified front. The only thing separating us from success (in academia, this translates into resources!) is cohesion. - Cohesion in terms of understanding that the academic standards for computing research can be asymptotic to those of engineering, the fine arts, the social sciences, and professional schools for some researchers, and more like mathematics and the hard sciences for others. (How often we shoot ourselves in the foot over this - particularly at times of promotion and tenure). - Cohesion in terms of presenting ourselves to the world as a profession as Denning suggests, rather than as an academic discipline. And, perhaps most important, cohesion in terms of representing ourselves as first and foremost indispensable problem solvers rather than as technicians or theorists.

    We are in the business of enabling innovation and creativity - and that will prove to be an easy concept to sell to politicians, executives, managers and administrators alike.

    REFERENCES:
    [1] Bryant, R. and M. Irwin, "Current and Future PhD Output Will Not Satisfy Demand For Faculty," (the 1999-2000 Taulbee Survey), Computing Research News, March, 2001, pp. 5-11.
    [2] Computing the Future, Computer Science and Telecommunications Board, National Research Council (1992).
    [3] "Academic Careers for Experimental Computer Scientists and Engineers." Computer Science and Telecommunications Board, National Research Council (1994).
    [4] "Information Needs in the Sciences: An Assessment, " Research Libraries Group, (1991).
    [5] Denning, P., "Crossing the Chasm," Communications of the ACM, 44:4, pp. 21-25.