5
Training, Retraining, and More Retraining

This chapter discusses the implications of likely shifts in demand for computing professionals and of current trends in their supply for education and training programs. It builds on discussions in earlier chapters about continuing shifts in the nature and mix of computing professional jobs and the need for a strong basic education that will prepare individuals for further training over time.

OVERVIEW

A principal mechanism for adjusting the supply of people to better fit job opportunities is investment in education and training. As discussed in Chapter 3, the number and mix of job opportunities for computer specialists are changing; as discussed in Chapter 4, education relevant to computer specialists is available in a wide variety of programs, institutions, and delivery mechanisms. Not surprisingly, discussions at the workshop revealed differences in perspectives between educators (primarily computer scientists) and employers in industry, but they served to highlight the principal issues for education and training of computing professionals.

In introducing the discussion of education and training at the workshop, Linda Pierce voiced the goal of reaching "a consensus around a core set of knowledge and skills" to support future career paths. Discussions at the workshop suggested that although there



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 78
Computing Professionals: Changing Needs for the 1990s 5 Training, Retraining, and More Retraining This chapter discusses the implications of likely shifts in demand for computing professionals and of current trends in their supply for education and training programs. It builds on discussions in earlier chapters about continuing shifts in the nature and mix of computing professional jobs and the need for a strong basic education that will prepare individuals for further training over time. OVERVIEW A principal mechanism for adjusting the supply of people to better fit job opportunities is investment in education and training. As discussed in Chapter 3, the number and mix of job opportunities for computer specialists are changing; as discussed in Chapter 4, education relevant to computer specialists is available in a wide variety of programs, institutions, and delivery mechanisms. Not surprisingly, discussions at the workshop revealed differences in perspectives between educators (primarily computer scientists) and employers in industry, but they served to highlight the principal issues for education and training of computing professionals. In introducing the discussion of education and training at the workshop, Linda Pierce voiced the goal of reaching "a consensus around a core set of knowledge and skills" to support future career paths. Discussions at the workshop suggested that although there

OCR for page 78
Computing Professionals: Changing Needs for the 1990s were areas of agreement, the community lacks consensus regarding career paths and core knowledge or skills, in part because of the widely shared view that the field—and therefore any related career path—is so dynamic. Workshop discussions centered on two themes: the appropriateness of the education of entry-level personnel, and the need for periodic retraining to keep personnel in the labor force up to date. Pierce spoke for many when she remarked that, with technology changing in cycles as short as two and a half years, ''the ability to forecast and prepare for this future skill mix will certainly grow more complex in the future." William Gear sketched out how the evolution of the field may translate into rising educational requirements for those engaged in computer-based technology development: Down at the very lowest level, you might even find people with associate degrees helping some very small company in some minor way with modifications of spreadsheet software . . . . In the future we are going to . . . see those people go through a [4-year] degree program that is heavy in the application area, probably not a major in computer science or management information systems, whatever it is called, but a minor to develop the necessary computational computer science skills and a major in the primary area of application. Gear suggested that the trend toward embedding more and more computer devices or systems into other equipment or systems will result in significant new skill requirements and potential problems in meeting them. The unresolved question is whether skill needs will be met in the classroom or on the job. Education and training are not preordained outcomes; they present choices for multiple parties—individual students, educators, and employers—whose outlooks often differ considerably. Ian Rose articulated the tensions between the interests of the individual in career development planning (including personal choices in education, training, and jobs to nurture his or her own career as a computing professional) and the interests of the employing organization (for which the individual's needs are secondary) in skills assessment and succession planning. James Williams, chair of the Department of Information Science at the University of Pittsburgh, pointed to the need for career counseling: "[People] need some kind of help because they don't have time to pay attention to all the nuances of what is happening in industry. They are busy doing their work." Participants from industry echoed Williams' assessment that planning for both education and training involves shared responsibilities.

OCR for page 78
Computing Professionals: Changing Needs for the 1990s Yet several participants suggested that today's computing professionals are not always as flexible and able to move with the technology as employers would like. Thus the challenge is how best to prepare people, through education and training, for the kinds of change the field may require of its practitioners and to inculcate the capability to be adaptable in response to these changes. Part of the problem, which affects many scientific and technical occupations, may reflect fundamental difficulties in harmonizing technical and managerial perspectives. According to Robert Weatherall, the initial education and work experience of engineers, and, by extension, of other segments of the computing professional community, may sow the seeds of discontent among engineers. He cited studies about the frequency with which engineers quit the field and go into management, often via M.B.A. programs. Weatherall suggested that one way to alleviate this problem is for engineering programs to encompass management issues, thereby helping to prepare engineers to progress into management and helping them to understand the relationship of their technical work to the operation of the organization as a whole. Participants also noted that in some organizations, efforts have been made to develop attractive technical career paths that parallel managerial career paths, but existing efforts do not solve the larger problem. Alan Fechter speculated that making technical career paths more rewarding would become a more urgent need over the next 10 or 15 years "because it is going to be harder for firms to recruit at the bottom. So, [employers] are going to have to think more carefully about how to utilize the resources they already have in-house." EDUCATION CURRICULA Going hand in hand with a lack of consensus on how the field and its practitioners should be labeled (see Chapter 2) was a lack of consensus on how accreditation might be used as a means of standardizing computer science programs and on what the core requirements should be for such programs. With respect to accreditation and standardization, Marvin Zelkowitz contrasted the conditions in academic computer science with the more stable and predictable conditions in such fields as physics or chemistry: "When I am asked what is the core of computer science, my comment is that it is whatever the computer science faculty want to teach." Similarly, Peter Freeman argued that "all of us have a strong responsibility to continue to push the formation of a core computer science discipline, and that is only going to happen by the continued

OCR for page 78
Computing Professionals: Changing Needs for the 1990s healthy production of Ph.D.s, in some sense almost forcing them into the market." These observations were reinforced by the comments of Joseph Turner and others about the evolution of criteria for departmental accreditation. Although such criteria exist and seem to be becoming more widely adopted, accreditation is not a major driver for the toptier, Ph.D.-producing departments. One reason may be the current lack of consensus about the ideal content for the field; Zelkowitz, for example, complained that the accreditation criteria were too focused on basic computer science. Another concern expressed was that accreditation requirements do not keep pace with the fast-moving field. With respect to the nature of the core curriculum, workshop participants debated the degree to which programs should be vocationally oriented. Those discussions reflected the divergent perspectives that exist between individuals in the academic research community and those in industry. The major area of disagreement concerned how much colleges and universities can or should do to teach either work-process skills or specific techniques. Zelkowitz acknowledged that educators have had difficulty with applications-related material, although he cautioned against an overly vocational approach. He suggested that while computer science departments started out with a "practical" orientation, they were likely to become theoretical over time, for two reasons. First, individuals oriented more toward applied work are more mobile, because they have more job opportunities; second, universities are not conducive to the undertaking of lengthy, large-scale experimental projects. Freeman observed, "There are certain specialties—systems software, systems building more generally—[for which] we basically cannot find good people either in quality or quantity . . . . And, yet, those are precisely the areas in which industry wants the most training among our students." Speaking from industry, Robert Kraut remarked that "In Bellcore's research organization, there are a large number of Ph.D.s in computer science, but there are very few of the sort that Bellcore is looking for, which are the systems researchers." A shift toward theory would weaken the fit between the education supplied by colleges and universities and the skill sets preferred by industry. Zelkowitz and other participants also noted that universities traditionally have not devoted much effort to specific issues of concern to industry—project management, teamwork, cost estimation, evaluation of alternative designs, risk management, for example—because those issues are crowded out by other items in the curriculum and because typical computer science faculty are not sufficiently knowledgeable about them.

OCR for page 78
Computing Professionals: Changing Needs for the 1990s David Gries summed up the frustration of computer science faculty with requests from industry for more practical and/or specific training. He observed to the industrial attendees, "You people want computer science to do more and more"; and he noted that undergraduate programs have only so much capacity for content. Ultimately, workshop participants agreed that the principal function of colleges and universities is to provide a foundation (including concepts, problem-solving skills, and so on). As William Lupton explained: [T]his gets at the heart of what we in the universities are all about in the first place. We teach students how to think. We teach students to . . . solve problems, using the theories and principles of the discipline and apply them to different challenges in the workplace. If the challenge changes, then they apply those principles again. The jobs may change, the tasks may change, but the paradigms don't change that often. If [there are] requirements for a change of paradigm, then they can go back and get retrained for it. I think what is missing [from] this discussion . . . is that when we develop students, we have developed a person that thinks and . . . can apply that thought process to the particular job. If that job changes, the person should be able to reapply those principles again and still be effective. Gries argued that some of the limits on program content reflect fundamental faculty limitations: [O]ne of the problems that we face in the field is that over half of the people who are doing the teaching at this point in all the colleges and universities and community colleges do not have a computer science background themselves. That is a real problem because of all the significant advances that are happening these days. We have object-oriented programming, functional programming—[but] most of the people teaching don't know about these things. . . . [T]here is a real . . . need for continuing education for computer science teachers. Similarly, Gries pointed out that computer science departments are not likely to address organizational and management issues (e.g., cost-estimation and teamwork) through case studies because "that requires the teacher to know something about those topics and that is not the case for most computer science faculty members." Paul Young contended that, overall, universities are responsive, if not rapidly so. Freeman, in fact, addressed the dangers of too quick a response from academia: [A]s educators, we have a responsibility to lead, not just to follow, the marketplace demand . . . . [T]he danger is that if we listen only

OCR for page 78
Computing Professionals: Changing Needs for the 1990s to the marketplace, of course, what we hear is, "We want more people that know object-oriented programming." Well, that is great, but that is not what you are going to need 5 years from now. I think all of us here know that and none of us know what you are going to need in corporate America or industrial America 10 years from now. Yet, the product that we produce in universities presumably has a useful lifetime of 40 years. So what the universities must continue to do is to produce that long-range product. Freeman also stressed the need for more interdisciplinary education: "[W]e need a new educational product that has . . . combined . . . parts of [computer science, electrical engineering, and business/management] . . . . I look at that as a third generation of computer science, what we are calling computing, which is a combination of the core computer science and some other discipline." Among industrial participants, Don McLean argued that broader perspectives in general, drawing from liberal arts, would be valuable. Paul Maritz, acknowledging some of the limits of what academia can do in providing work-related training, suggested that more specific skills may be the province of master's and doctoral programs: The university can't give the student big-systems experience. They can't put together a 100-person project that runs for 3 years, which is what you have to do . . . . [I]ndustry has to supply that side of the picture. What you want [from] the schools is training in analytic thought, in good work habits, in curiosity. These are the types of things that you would like to see come [from] the undergraduate institutions . . . . [T]he postgraduate institution is where we can look for more specific skills, somebody who really has been trained in depth in computer graphics, for instance, or in computer linguistics. That is where you can look for people who understand modern principles and techniques that can be applied. Robert Weatherall observed that most of the communication between employers and universities comes from big companies, a factor that might bias the message. Finally, Leslie Vadasz urged greater and more thoughtful communication between industry and academia: Some of this discussion reminds me of problems in dealing with our customers. In a high-technology business, there is a difference between what a customer wants and what he needs. Customers generally can tell you their "wants" based on their experience. But new technology can create capabilities beyond their "wants," and you need to evaluate if they really ''need'' it. There is a similar interface between universities and employers. We in the industry can probably state our "wants" on the basis of our most recent or current

OCR for page 78
Computing Professionals: Changing Needs for the 1990s experiences. But can we define what future skills we need? Maybe yes; maybe no. I believe that it behooves the supplier—in this case the universities—to understand what future issues their customers—in this case the employers—will face, so that they can better supply what their customers really need. This is where industry-university dialogue has to continue. We don't go out to hire a Ph.D. As Paul [Maritz] said, we hire on the basis of certain accomplishments . . . . It is very difficult to specify demand in a very fast-moving field like this. TRAINING Training, per se, is largely a concern for computing professionals engaged in applications and systems development and possibly deployment. Training is generally provided or paid for by employers.1 In addition to on the job training, sources include workshops, seminars, conferences, and programs offered by universities and professional training organizations. Speaking from their experience, workshop participants described a labor market in which employers, especially large firms, are investing more and more in training people to meet their particular needs (Box 5.1); this practice allows them to focus on specific skills when making hiring decisions, as discussed in Chapter 3. Workshop participants agreed that training is an ongoing concern for computing professionals because of the dynamic evolution of computer science and technology. As a result, they suggested, computing professionals should be educated—and employed—with an orientation toward lifelong learning and periodic retraining. James Williams identified four principal categories for retraining: literacy or basic identification of things and processes; concepts, "where you need to understand some relationships and you need to understand how [something] fits into some structures or some forecast"; knowledge, "where you really need to understand how to solve problems, how to do analysis and synthesis and evaluation of a particular area"; and skills acquisition, the ''how-to" category. It was no surprise that the issue of how much retraining should be provided and how frequent retraining should be was not resolved by workshop participants. Barbara Wamsley suggested that training would be a constant, recurrent activity: "When you know you are dealing with an occupation with a half-life of 2 to 5 years, you are going to have to train these people in 2 years anyway if you are going to keep the same quality that you hired." Ian Rose noted that, along with changes in skill requirements

OCR for page 78
Computing Professionals: Changing Needs for the 1990s BOX 5.1 Employer-based Training "[Y]ou simply can't afford not to be in the internal training business. Whether internal training means having internal training staff or tapping the commercial suppliers or, indeed, the local colleges and universities, we simply have to be in the business of training, and I would suggest more and more as change increases."—Don McLean "It is interesting to me, sitting in a careers office, to see firms like Arthur Andersen, Andersen Consulting, Electronic Data Systems, the big brokerage firms like Morgan Stanley, and then the big insurance companies. . . . [T]hey seem to think that the thing to do is look for the bright folk and then have a very elaborate training program, so they will know exactly what they should be learning and doing within [for example] the insurance company."—Robert Weatherall "What we hear also from industry is that they think the universities should do what they do well, which is to provide more basic science and broader intellectual support, but not try to teach people how to do a very specific product-oriented task, which might well turn out to be out of date. That it is an appropriate division of labor: that universities should do education, and that companies should then do the training."—Tora Bikson associated with changes in technology, concurrent changes are occurring in organizational structure (e.g., organizational flattening, outsourcing, or downsizing) that are altering the opportunities for advancement and professional growth among personnel in industry. Linda Pierce also enumerated managerial concerns that complement concerns about technological change: product delivery needs, business savvy, ease of use of technology and applications, ability to work in teams, more formality in product research and development, higher-quality expectations, and so on. These concerns should motivate new directions in education and training. Rose pointed out that individuals may not keep pace with how technological and organizational shifts affect advancement opportunities. Thus, they may develop unrealistic aspirations. At the same time, contended Peter Freeman, managers may not be sufficiently knowledgeable about the options and implications of new technology to make the best training decisions about either technology or jobs.

OCR for page 78
Computing Professionals: Changing Needs for the 1990s Both individuals and managers are struggling to forecast needs and opportunities in areas that involve computing; the uncertainty and dynamism of the time imply that flexibility and an ability to shift to new skills and responsibilities are especially valuable. Pierce presented the employer's perspective on the need for flexibility: "[O]ne thing we have to expect is that these people understand that they have chosen a field that is going to have rapid change in it . . . and they have to be committed to learning as a lifelong thing, if they are part of this kind of work." Participants from industry acknowledged that some people would feel more pressure to change in their skills than others. This variation in the training needs of industrial employees reflects, in part, the need to support or maintain systems throughout their life cycles. For example, Chris Caren contended, I don't [believe] that everyone is involved in this change process. We've got systems that we have had for 20 years that are written in [the 1950s-vintage programming language] COBOL. They will probably be around for another 20 or 30 years if we can't afford to change to anything else. . . . [P]eople that are involved in the maintenance of those [systems] . . .—a vast number of people involved in our computer applications—hopefully are going to get some retraining. But they wouldn't need to have retraining. Then there are others in the very mobile world of new projects . . . [who must] have the retooling, the reskilling, and so on. That is where you put your more flexible people. So, this renewal process probably only really needs to be done for a percentage, and I don't know what it is—30 percent, 40 percent of your work force. Workshop participants acknowledged that individuals differed in their receptiveness toward retraining. Williams noted that incentives, including financial support or promotional opportunities, can help to encourage retraining. John McSorley described how some 70 percent of those whose skills are no longer needed at Apple are unwilling or unable to be retrained, and as a result they leave the company. Shelby Stewman, professor of sociology and demography at Carnegie Mellon University, speculated that applications skills are neither as transferable nor as malleable as most might expect. He observed that workshop participants and people he has interviewed in industry have indicated that it is difficult to transfer people across system and application areas; it is often easier to hire someone new with the appropriate skills. Participants were, in the one and one-half days of the workshop, unable to articulate more than a general sense that needs for frequent retraining will continue in the future and may even expand, given expected trends in new technologies.

OCR for page 78
Computing Professionals: Changing Needs for the 1990s ISSUES AND CONCLUSIONS Changes in skill requirements will be a way of life for computing professionals. This conclusion arose from discussions of education and training as well as supply and demand. Frequent retraining may be necessary to meet shifting needs for skills. The educational system can support this requirement both by producing graduates with broad and flexible skills and by contributing to more effective and more available retraining through interactions with industry. Corporations must reconcile themselves to the financial and organizational commitments implied by continuous shifts in skill requirements. Changes in skill requirements may result in some dislocation for individual employees. But rather than simply replacing employees, in the long run it may be more effective to update their technical skills while leveraging all of the other skills and capabilities required of employees to make organizations effective. NOTE 1.   As discussed in Chapter 4, some entry-level training is also provided in 2-year and other short-term educational institutions, but those programs generally are oriented toward computer users or associated production and clerical personnel rather than computing professionals.