The History of Computer-Assisted Career Guidance Programs (CACGSs)

home > essays > CACGSs
A History of Computer-Assisted Career Guidance Systems
by Andrew D. Carson, Ph.D.

There have been several efforts to trace the history of computer assisted career guidance systems (CACGSs). Among the earliest, the proceedings of the "Eighth Invitational Conference on Systems Under Construction in Career Education and Development" (Career Education and the Technology of Career Development, 1971) included several reports by researchers of their development of programs. Fifteen years later, Sampson, Shahnasarian, and Maddox (1984, pp. 3-6) brought the history up to date. Rayman's (1990) excellent chapter first proposed the concept of generations of CACGSs, outlined below. Katz's (1993) book on the history, rationale, and future directions of CACGSs is the most thorough treatment of the problem up through the fourth-generation systems. Carson and Cartwright's (1997) article extended Rayman's generations-based approach to cover networked computers. Most recently, Reile and Harris-Bowlsbey (2000) provided a brief history of the use of computers to support career planning.

What is interesting about these various histories is that they tend to ignore the broader historical context in which such developments occurred. Our overview of the history of CACGSs assumes three main phases: conceptual and early research (before the end of World War II); command CACGSs created during the Cold War; and systems developed since then, and largely spurred by the growth of the World Wide Web and the Internet.

Phases in the History of CACGSs

I. Conceptual and early research. Applied psychologists have long sought to create technologies that promise to reduce the costs of administering and interpreting vocational tests and inventories. Clark Hull (1928) proposed a calculating machine that could both administer aptitude tests and from the results make career recommendations. Although computers were used to score occupational tests as early as 1940 (see Flanagan, 1961), Hull's conceptual innovations could not be implemented until the growing availability of electronic general-purpose computers in the 1950s.

II. Command CACGSs. Enter the Cold War. The National Defense Education Act (NDEA) in 1958 (spurred in large measure by the American reaction to the launch of Sputnik), together with the Vocational Education Act (VEA) in 1963 (also closely tied to the American Cold War agenda), provided substantial investments in innovation in career guidance research and the development of systems, a good portion of which flowed in the direction of computer-based methods. Thus, the impetus to build effective computer-based career guidance systems appears to have been related by the desire to catch up with the Soviets, who were perceived as having taken the lead in directing qualified youth into careers in science, math, engineering, and other technical fields. The American Institutes for Research (again, closely connected to supporting Cold War initiatives) continued to fund research and conferences on the topic at least into the 1970s, by which time a number of research programs -- including those creating SIGI and the forerunner of DISCOVER -- had begun to develop both prototypes and practical applications for computer assisted career guidance systems (CACGSs). Since then, several individuals and organizations have built on and refined these early programs, all essentially spin-offs of Cold War research and development. Student or client usage of command CACGSs was generally sold not to individuals, but rather to institutional customers such as public school systems. Essentially, the Armed Services Vocational Aptitude Battery (ASVAB) became the largest and most heavily funded command CACGS, with funding directly from the United States military and with usage in at least 80% of United States schools at its peak.

The technological development of command CACGSs has advanced through four generations (see Rayman, 1990): batch processing, online systems for information retrieval, systems in support of career development, and microcomputer-based systems (See Table 1). However, each new technology supporting command CACGSs does not so much replace the earlier developed ones as much as just become added to the mix.

1 Batch processing, 1950s onward. Mainframe in one place, you send materials (say, inventories completed by clients) to it for processing. Clients get back reports and information.

2 Online systems for information retrieval, 1960s onward. Mainframe in one place, you connect to it (say, by phone line) and interact with it to retrieve information.

3 Systems in support of career development, late 1960s onward. This is a direct extension of second-generation, adding new capabilities that allow the computer to simulate a system of career exploration. Examples: SIGI and DISCOVER, and these systems move to near duopoly.

4 The microcomputer, 1970s onward (or PC, as we would call it now. Fourth-generation systems migrate to microcomputers (generically called PCs today). Several new smaller systems or modules develop and compete; the ETS/ACT duopoly begins to crack.

III. Free CACGSs. The end of the Cold War and the victory of the Western democracies occurred about the same time as the birth of the World Wide Web and the growing use of the Internet as a medium for CACGSs. Carson and Cartwright (1997) argued that CACGSs had now entered their fifth-generation through making use of networked and Web-based computer systems. Where there were relatively few CACGSs created in generations one through four, there have been many developed in generation five, thus representing the major growth area in career assessment (see Oliver & Zack, 1999). Most of these CACGSs are free of charge and most emphasize the essential "freedom" that individuals have in making choices; for example, gone is the government-subsidized encouragement to enter occupations considered vital to the national defense effort.

Unfortunately, the documentation of the practical efficacy of CACGSs - and especially the Web-based ones - lags far behind their conceptual and presumably commercial potential (see Hinkelman & Luzzo, 1997; Oliver & Zack). In short, everyone has been so busy creating and trying to make money (or at least not to lose money) from CACGSs that there has been little time or money left over to show that they actually work. In addition, traditional providers of command CACGSs have faced great difficulties in deciding how best to cope with and adapt to the Web and the Internet.

Some Trends that Will Affect CACGSs

It is mandatory in essays such as this to turn from a discussion of history to predictions of future trends. As Oliver and Whiston have already identified a good set of trends (proliferation of career assessment websites, increasing research on sites that are free or charge, and the growth of career outcome assessment on Internet-based systems), we shall identify new ones. We believe that many of the trends that will affect the future direction of CACGSs lie beyond the traditional scope of expertise of counseling and vocational psychologists.

Trend 1: New Players Develop CACGSs. CACGSs on networked computers disrupt the traditional CACGS market. Just about everyone engaged in developing Web-based business strategies seems to have read Christensen's (1997) The Innovator's Dilemma, which provides a theory why truly innovative technologies often disrupt rather than aid established companies. It appears to us that the Web is disruptive of the traditional business models of publishers of third and fourth-generation CACGSs for three main reasons: consumers will indeed opt for lower quality but less expensive CACGS almost every time, and fifth-generation systems are the low-cost leader; fifth-generation systems present more confidentiality and security issues (especially a problem for skill tests), and traditional companies are loathe to give up control and especially the security of test items; and fifth-generation systems don't cost much to create (lower barriers too entry). The more you spend (money and time) on creating high quality fifth generation CACGSs, and the more effort you spend in transporting all of the features from older ways of doing things onto the new Web-based platform, the greater the risk of falling behind fleeter competition that nimbly develops products that, although of lower initial quality, do work, and reach market first.

Now, if consumers really do demand the features (e.g., known brand, higher reliability and better validity indicators), then the new, disruptive entrants will fail, and the slower, more stable, older players will succeed. In the broader world of e-commerce, no one is certain how the balance will tip overall, but current impressions are that it varies by industry. The experience thus far in career interventions and CACGSs is that the Web-based systems are disruptive, that new entrants will probably dominate, and that the traditional players either better rapidly get their act together or risk dying. Part of "getting their act together" may be to not spend overly excessively on traditional research to validate the sites, if in so doing they only help to drive themselves out of business. Oliver and Chartrand (2000) appear to recognize this in their acknowledgment that funding concerns may make developers reluctant to spend scarce funds on validation research, but their only solution is to suggest that the developer raise funds for such research through an initial public offering (IPO), not a very promising strategy for those who have followed the news around tech IPOs over the past year.

A more complete history of this situation would compare the fortunes of several types of companies and how they develop CACGSs on the Web, such as:

the fourth-generation leader (say, ACT with DISCOVER)
a test publisher with some fourth generation test foothold expanding their system on the Web, but not willing to cannibalize their paper-based or fourth generation products in doing so (say, CPP or PAR)
a test publisher or consulting firm with no previous CACGSs products placing (or considering placing) a big bet on the Web (say, The Ball Foundation or Haldane & Associates)
a media or education company with significant Web experience expanding into the development of a CACGS (say, the Wall Street Journal or AOL Time-Warner)
a dotcom already successful in job search services expanding their line into a full CACGS service (say, Monster.com or HotJobs.com); and, finally,
a garage-based dotcom operating on a shoestring and on weekends, possibly with an unpaid staff of one, and cobbling a full CACGS together through Web-enabled partnerships. Which option should win out in the end? If the Web is truly disruptive, the options toward the end of the list should fare better - and ultimately develop better systems -than those listed earlier.

Trend 2: More Usability Research . Usability research is replacing traditional psychometric and outcome research with CACGSs. Usability research increasingly plays a major role in Web development efforts, and presumably CACGSs will not be immune to this. Web usability research has emerged out of the fusion of three disciplines: human factors engineering, cognitive science, and Web design. Should we push for traditional outcome validation research for CACGSs, or should we first advocate for effective usability research? Oliver and Whiston's (2000) hopes for an increase in career outcomes research for fifth-generation systems to the contrary, we suggest that usability research may be the most practical use of research budgets for fifth generation CACGSs at this stage of their development. You want to know how well interfaces work, and you need the information immediately so you can fix the system. Prince and Chartrand's (2000) discussion of the development of CPP's CareerHub site contains discussion of the importance they placed on pilot testing; this suggests they placed special importance on usability. CACGSs on the Web do not need product cycles of a decade, say, as with the Strong Interest Inventory. For example, product cycles on the Web can be by the quarter for major system enhancements, and daily for minor ones. Usability research is more useful and more appropriate, given this rapid churn. Essentially, it is more ongoing evaluation research (in the form of usability) that you need with CACGSs than validation research of fixed interventions.

The most widely read work on the subject of Web usability is Nielsen's (2000) Designing Web Usability, but also see Forsythe, Grose, and Ratner (1998). Also, an excellent site for reviewing current developments in Web usability is UseIt.com. See especially that site's "Ten Usability Heuristics" (2000), addressing heuristics also applicable to the design of CACGSs: visibility of system status; match between system and the real world; user control and freedom; consistency and standards; error prevention; recognition rather than recall; flexibility and efficiency of use; aesthetic and minimalist design; help users recognize, diagnose, and recover from errors; and help and documentation.

Trend 3: XML, Hybridization, and Cobranding. CACGSs are, along with all other major data-intensive applications delivered through networked computers, increasingly designed to separate content from style. This has had a number of interesting effects, including making it more difficult for companies to tightly control the branding and trademarked aspects of their products. This trend is obvious in the rapid move in application development away from hypertext markup language (HTML) and toward extensible markup language (XML). Companies that rely on XML for application development generally rely heavily on created well-documented structures to code their content into; style and presentation issues are handled separately. This becomes very useful, for example, when you want to publish content across a variety of platforms, say, desktop computers, personal digital assistants, telephones (yes, you can now get websites by phone), and, in the not too distant future, through electronic paper and other media now in development.

When you separate content from style using standard coding methods, it allows companies and individuals to trade such documents easily among themselves, and to present them through any available site or other interface. This leads to a lot of syndication, not only of content, but also, essentially, of interface and style standards. If you want to see an example of trading styles, see some of the various "skins" available though the music sites on the Web.

At any rate, this trend makes it difficult to imagine a purely ACT or ETS or PAR or Chronicle Guidance or even O*Net CACGSs. Instead, look to the leading Websites to see what's coming. You might see "DISCOVER, powered by Chronicle Guidance" for occupational information, or CareerPath assessment powered by CPP, Inc.. Imagine going to Monster.com for matches of all jobs matching your Holland types, and a personalized results page incorporating not only job leads but also descriptions of RIASEC-oriented descriptions and "powered by PAR."

Trend 4: More Profits; Nonprofit Problems. CACGSs, and career intervention more generally, becomes very profitable for the providers that survive; that's due to the great economy of scale involved. It's also the business to business contracts afforded by selling to states, or school districts, or subscription websites that serve as content integrators.

A related issue is the increasing difficulty that nonprofits with successful career-related websites have in maintaining that status, when the profit motive becomes so tempting. The Internal Revenue Service (IRS) is increasingly looking at so-called nonprofits to see if they really are. Many nonprofits are starting to make a lot of money from their Web-based operations and vocational-oriented consulting operations. It's one thing when you are a nonprofit that can only reach one city. It's quite another when you reach the world and have a significant tax break. Competitors complain! ETS has already had to spin off the Chauncey Group. ACT presumably faces similar challenges. At any rate, expect to see many nonprofits finally accepting reality and going for profit, and when that happens they may find they like it, and that they grow fast.

Trend 5: Stronger Copyright Protections. Weakening of fair-use provisions of the copyright law, coupled with increased corporate litigation to protect copyrights) will make it more difficult to legally take advantage of the wealth of content available over the Web and other networked computers, especially for for-profit organizations. Far from being a new land of unfettered freedom from traditional regulations--and especially copyright restrictions--the Web and the Internet are shaping up to be a tightly regulated space in which "code" becomes closely monitored against rule-breaking (see Lessig, 1999). Although fair-use provisions have traditionally shielded nonprofits (especially those serving educational roles) against charges of copyright infringement, aggressive defense of copyrights against even nonprofits has eroded such defenses, especially when nonprofits offer their services world-wide over the Web.

Tax authorities (such as the IRS) are also reviewing with increased scrutiny those nonprofits against which others lodge complaints related to unfair competitive practices; infringement of copyright might spur such complaints and reviews.

CACGSs are especially prone to violate copyright claims because their creators may seek to integrate material from a variety of sources, in the course of which copyright violations may occur. Even so-called deep-linking into pages of websites below the "homepage"--a very common practice with CACGSs -- might be considered a violation of copyright unless express permission is provided by the holder of such rights; copyright law on this point is not set in stone.

Trend 6: Growth of Open-Source. A major trend in software and content development in recent years has been the open-source movement. Trend five increases the risk of using, without permission, the copyrighted materials of others. In response to such assertion of copyrights with software has emerged the open-source movement, created by those who want to maintain free access to software and the ability to modify it freely. The open-source movement works, as witnessed by the success of Linux, a serious competitor to Microsoft's PC operating systems. In open-source programs, anyone may study the source code of a program, may modify it, and may redistribute it, generally under license from the original creator. Open-source systems often wind up being more robust and bug-free than their commercial counterparts, mainly because of their openness to admitting imperfections in the software, and their reliance on a community of bright collaborators to fix such problems and make the software work.

As counseling and vocational psychologists, we might consider banding together to create our own open-source CACGS platform, essentially the Linux of CACGSs. We shall save further exploration of this idea for another setting, but we suggest it is an exciting idea and one well-worth discussing. It is also likely that at least one provider of an existing commercial CACGS may consider shifting their system to an open-source format in order to lock-out competition, largely eliminate their development costs, and set the stage for a shift in their business focus to the delivery of consulting and training functions to users of the thereafter open-source CACGS.

Some Warnings

Carson and Cartwright (1997) expressed concerns about the possible negative consequences of fifth generation CACGSs. It is our belief that some of these were not only correct but are even greater threats today; the two threats that have increased the most over the past few years are the risks of job loss or deprofessionalization among counselors, and loss of privacy.

Although the promise of Web-based and other fifth-generation CACGSs may offer wealth or at least much improved earnings potential to a handful of counseling and vocational psychologists, it may also do so at the risk of moving them further and further from their person-to-person service roots, and also at the risk of increased anxiety due either to shifting to contractor status (quarter-to-quarter or even month-to-month jobs) or to working for start-ups with uncertain futures. Vocational and counseling psychologists considering making the leap to new economy career tracks should realize that the increased rewards offered by that route may come at a high price in terms of anxiety, long-hours, and job churn (see Reich, 2001). Moreover, CACGSs and traditional counselors may be players in the same zero-sum game; the creation of really effective CACGSs may tend to deprofessionalize an entire class of career counselors into becoming little more than tenders of the machine, or even leaving them out of the system entirely. Although this issue did not appear first with fifth-generation CACGSs, it has grown in importance in recent years.

Concerns related to privacy and confidentiality have also increased in the past few years (see Sampson & Lumsden, 2000), although improved encryption methods may reduce the chance of unauthorized snooping. More at risk are the datafiles maintained by the providers of CACGSs, and the possibility that they might be used for other than counseling purposes. (Of course, clients might not mind such repurposing of their confidential information, but that's another issue.) A more thorough exploration of these issues lies beyond the scope of this paper but is worthy of attention.

Conclusions

In summary, computer-assisted career guidance systems (CACGSs) are hardly a recent invention, as they have evolved across nearly four decades of practical development, and over almost eight decades (since Hull) of conceptual selection and winnowing. Although there is now enough history behind CACGSs to make its recounting interesting (perhaps especially in the transition from command to free CACGSs), it remains a subject in which we have much more interest--and perhaps even anxiety--in looking forward than in looking back.

Links

SIGI Plus homepage (ETS).

DISCOVER (ACT).

Clark Hull (1884-1952), psychologist who predicted machines that could translate test scores into occupational recommendations.

Main page for Lawrence Lessig's Code: And Laws of Cyberspace.

XML.com, central page for information about Extensible Markup Language (XML).

Jakob Neilsen's homepage for usability information (useit.com)

National Defense Education Act (NDEA)of 1958; summary of major provisions.

References