All members of the scanning committee should become more aware of their ongoing passive scanning. The special screen of the scanning criteria should be added to the flow of each person's continuing resources; it Is a level of sensitivity that has to be learned with experience. It must be a rule of the committee that information in any form is acceptable. The process of passing notes, clippings, or copies from any resource must become second nature. The scanning coordinator or staff person will have the responsibility to process the incoming flow for the committee's formal review.

The committee must now address the question of the resources it will actively scan, and it must consider several aspects of the available resources in making the decision. First, a survey of the committee will show the specific resources included in its passive scanning. Then the committee must determine the kinds of resources it should be scanning, which involves the content and the kind of research-for example, germane to all issues, germane only to special issues, emerging or first impression of issues, the spread of issues.

In the process of assigning resources to issues, the committee should also address the question of the mix of the media it is using-from periodical to annual publications, from print to electronic forms-and it should review its resources to determine a balance in the mix of the media. A list of journals focusing on the general field of higher education or on specific aspects of the field is shown in Appendix A, and Appendix B includes publications focusing on external issues.

Popular scanning resources. Newspapers are a major scanning resource, and the members of the committee should cover four to six national newspapers to balance the newspapers' particular focuses and biases: the Nevi, York Times for its focus on international affairs, the Washington Post or Times for their focus on domestic political developments, the Chicago Tribune for its focus on the Midwest, the Los Angeles Times for its West Coast perspective, and one of the major papers of the Sunbelt. USA Today and the Wall Street Journal, with their emphasis on trends and forces for change, are perhaps the most popular newspapers of scanners. The national perspective should be supported by a review of the relevant major state, regional, and local newspapers.

Magazines, periodicals, newsletters, and specialized newspapers in each of the four major areas--social, technological, economic, and legislative/regulatory--should be included. But it is also important to include publications of special interest groups that are attempting to put their issues on the national agenda (Congresswatch, Fusion, the Union of Concerned Scientists, the Sierra Club, the National Organization for Women, and Eagle Forum, for example) and journals reporting new developments, such as the Swedish Journal of Social Change and Psychology Today. Although the list of scanning resources may appear formidable, the number of new periodicals added to existing resources may be quite small, for at most universities, some member of the faculty already sees one of the resources or it already is received in a campus library.

A special effort should be made to seek publications of the fringe literature-the underground press-as exemplified by the Village Voi(-e and other non-establishment publications. Depending upon the results of the survey of literature already being covered by members of the scanning committee, a special effort could be made to include publications like Ms., Glamour, Working Woman, Working Mother, Family Today, and Ladies Home Journal. Finally, the scanning literature should include a few wild cards--High Times, Heavy Metal, Mother Jones, for example. The scanning staffer should maintain a list of publications that are being scanned and the committee members responsible for scanning them. Ideally, each member of the committee should be responsible for three to four titles.

Additional resources for scanning include trade and professional publications, association newsletters, conference schedules showing topics being addressed and considered, and, in particular, publications of societies and associations involved with education and training. For example, many instructional innovations are surfacing in corporate training programs and are being discussed at annual meetings of the American Society for Training and Development and in trade publications like Journal of Training and Development and Training: The Magazine of Human Resources Development. As a further example, the forecasting movement and the concept of strategic planning developed in the business sector years before most individuals in higher education were aware of them as potentially affecting colleges and universities. Other industries--health care and social services, for example--may experience issues before higher education. Strategies for cost containment in the health care sector, for example, may well merit adaptation by higher education as funding support lessens (Morgan 1983).

A number of associations and societies track or advocate social change. The World Future Society, for example, publishes The Futurist, The Futures Research Quarterly, and Future Survey all of which are dedicated to the exploration and discussion of ideas about the future. The American Council of Life Insurance in Washington, D.C., publishes a newsletter, Straws in the Wind, and periodic reports on emerging issues called The ACLI Trend Report. In addition, major corporations use commercial services to supplement their scanning functions: Yankelovitch's Corporate Priorities, the Policy Analysis Company's CongresScan and Issue Paks, the Naisbitt Group's Trend Report, SRI International's Scan, and the Institute for Future Systems Research's Trend Digest. The more expensive outside resources are beyond the budgets of most colleges and universities and are not without their own liabilities (many of them attempt to cover all issues from all perspectives, making their results too general to meet the needs of specific organizations), and an overemphasis on outside resources violates an organizational requirement that the scanning function be developed within the existing structure rather than added on from the outside.

The scanning committee should make a special effort to include within the scanning process whatever fugitive literature it is able to obtain, that is, sources that are published privately and are available only if their existence is known and they are hunted down. Such literature would include, for example, the more than 25 articles, pamphlets, and other private publications now available on the new field of issues management compiled by the Issues Management Association in Washington, private publications on changing social values such as the 1981 Connecticut Mutual Life Insurance study, AT&T's Context of Legislation, and the publications of research organizations like the Rand Corporation, SRI International, or the Center for Futures Research at the University of Southern California. Fugitive literature often enters the established literature, but sometimes years after its initial private publication. Thus, it is necessary to develop personal and professional contacts throughout the scanning network to gain access to these materials. Professional associations like The World Future Society, the Issues Management Association, or the North American Society for Corporate Planning and their conferences can be major sources for fugitive literature.

Other resources. The scanning committee should tap the resources of its resident experts (Renfro and Morrison 1983b), best accomplished by the publication of a weekly or monthly scanning newsletter prepared by the committee's staff. This brief newsletter might present two to five of the more significant items recently found by the scanning committee. Such newsletters continue to build a constituency for the scanning process and an informal network for the recognition and appreciation of the results of scanning. The newsletter might be sent, for example, to all department chairs with an open invitation for their comments and for suggestions of new ideas they see in their fields. Colleges and universities are in a unique position to conduct scanning: Many organizations do not have the in-house experience that is available on most faculties.

Internal scanning newsletters frequently use political and issue cartoons found in major newspapers and in national magazines like The New Yorker. Such cartoons provide an important signal that at least the editors believe the issue has reached national standing and that some consensus on the issue exists for the cartoonist to create the foil and hence the humor. These cartoons serve the additional functions of communicating a tremendous amount of information in a very small space-a picture is still worth a thousand words.

After operating for a year, the scanning committee needs to review the clippings and articles collected and eliminate outdated materials. A staff person should have the responsibility of maintaining the files, opening and closing categories only with the approval of the whole committee. To keep the scanning from becoming outdated, the list of publications scanned should be reviewed and those resources that yielded little information in the preceding year dropped.

Operating an environmental scanning process requires a commitment of time and resources. It may be desirable for colleges to form consortia to share resources, following the example of the life insurance industry. Or they may develop cooperative arrangements with local corporations through which they receive scanning information, particularly projections of the region's economy and emerging technology. It is imperative, however, to establish an effective scanning system in this fast-changing world to identify as early as possible those emerging trends and issues that may so dramatically affect the organization's future.

Collecting judgments on an event's probability, impact, and degree of control can be done by using simple questionnaires or interviews and quantifying participants' opinions using various scales (for example, probability can range from 0 to 100, impact from 0 to 10). When all participants have made their forecasts, the next step is to calculate a group average or median score. Quantification is useful because it is fast, and it tends to focus the attention Futures Research and the Strategic Planning Process of the group on the subject rather than the source of the estimates.

Real surprises usually come from failing to keep track of small-scale developments in the external environment, not from excluding small-scale developments within one's own system. By systematically following these external developments it is possible not only to anticipate the directions and potential impacts of the slower, more pronounced, more profoundly influential changes but also to obtain the early warning needed for timely adjustments of strategy. Emerging patterns of events, the ebb and flow of particular sets of issues that can be revealed by close monitoring, provide a basis for forecasts relevant to policy. These forecasts are intuitive, to be sure, and perhaps seen only dimly in outline, but they are nonetheless the best forecasts available.

Even when the output from scanning consists of forecasts, we must still make our own judgments about the future, because we must decide what is relevant and we must make judgments as to whether we agree with the given forecasts. The same process is at play when we read newspapers, journals, reports, and government documents or listen to a broadcast. We constantly make personal forecasts on the basis of sparse and fragmented historical data in an attempt to distill the future that may be implied.

This process of trying to infer the future by mentally extending current or historical data is sometimes called "implicit forecasting." Such forecasting is obviously as useful as it is unavoidable when it comes to obtaining an appreciation of the broad outlines of possible futures. By itself, however, implicit forecasting is not sufficient when it comes to making today's decisions about our own most important long-range issues-the direction of a career, the development of a profession, the survival of an institution, department, or program, for example. In such cases, the need is also for methods that deal much more formally, systematically, and comprehensively with the nature and likely dynamics of future events, trends, and policy choices.

It is easy to see why our implicit forecasts of the general context are progressively less trustworthy as the questions at stake become more important. These forecasts are entirely subjective, they are no doubt idiosyncratic, they are often made on topics we are unqualified to assess because of a lack of relevant experience or knowledge, they rest very largely on unspoken arguments from historical precedent or analogy, and they are haphazard in that they are made primarily in response to information we receive that is itself usually developed haphazardly or opportunistically.

As futures research has developed since the mid-1960s, much work has gone into the invention and application of techniques intended to overcome these and other limitations of widely practiced methods of forecasting. In general, the newer methods are alike in that they tend to deal as explicitly and systematically as possible with the various elements of alternative futures, the aim being to provide the wherewithal for users to retrace the steps taken. The following paragraphs highlight some of these methods.

Genius forecasting
Apart from implicit forecasting, the most common approach to forecasting throughout history has been for a single individual simply to make explicit guesstimates about the future. In their weaker moments, many bright and otherwise well-informed people-including even futures researchers-are sometimes cajoled into offering such guesstimates, which typically take the form of one-line forecasts ("cancer will be cured," "no ship will ever be sunk by a bomb," or "the end is near"). But if they are persuaded to reflect on the future in a widely ranging way, to try to articulate the underlying logic of affairs and its likely evolution over time, to reason through the obvious alternatives and imagine the not so obvious ones, when in short they offer a careful but creative image of the future in its richness and complexity, then a much different process is involved. It has no common name, but in futures research it is often lightly called "genius forecasting." It is a powerful and highly cost-effective way to obtain forecasts if the "genius" is indeed thoughtful, imaginative, and well read in many areas.

The disadvantages of genius forecasting are clear enough to require no enumeration here. "In the end, genius forecasting depends on more than the genius of the forecaster; it depends on luck and insight. There may be many geniuses whose forecasts are made with full measure of both, but it is nearly impossible to recognize them a priori, and this, of course, is the weakness of the method" (Gordon 1972, p. 167).

If used properly, however, the strengths of the method usually outweigh its weaknesses. The probability of the integrated forecast produced by the "genius" is certain to be virtually zero. Time will show that the forecast was oversimplified, led astray by biases, and ignorant of critical possibilities. Yet the genius has the ability to identify unprecedented future events, to imagine current policies that might be abandoned, to assess the interplay of trends and future events in a far more meaningful way than any existing model can, to trace out the significance of this interplay, to identify opportunities for action that no one else might ever see, and to explain assumptions and reasoning. Although the genius forecast will be both "wrong" and incomplete, it will nevertheless have provided something very useful: an intelligent base case.

Occasionally, genius forecasts can serve as the only forecasts in a study. This approach makes excellent sense in studies being accomplished under severely constrained time and resources. Increasingly in futures research, however, studies are begun by commissioning one or more genius forecasts, which take the form of essays or scenarios of one sort or another. With them in hand, the investigators explore them carefully for omissions and inconsistencies, and then the forecasts are carefully pulled apart to identify the specific trends, events, and policies that appear to warrant detailed evaluation; that is, the most uncertain, problematical, intractable, and potentially valuable statements about the future can be selected. Being able to launch a more sophisticated forecasting effort from such a basis is much better than having random thoughts and blank paper.

Extrapolation of mathematical trends
Most forecasters and some practitioners of futures research use techniques of mathematical trend extrapolation that are well understood, rest on a fairly adequate theoretical foundation, convey the impression of being scientific and objective, and in skilled hands are usually quick and inexpensive to use. One of the most commonly used techniques is regression analysis, one purpose of which is to estimate the predicted values of a trend (the dependent variable) from observed values of other trends (the independent variables). Hierarchical regression models are sometimes referred to as "causal" models if an observed statistical relationship exists between the independent and dependent variables, if the independent variables occur before the dependent variable, and if one can develop a reasonable explanation for the causal relationship. A forecast of the independent variables makes possible a forecast of the dependent ones to which they are statistically linked, whether the case is simple or complex. In either case, however, the purpose behind causal regression models is always to explain complex dynamic trends (for example, college and university enrollment patterns) in terms of elementary stable trends (for example, demographics or government spending).

When cause is not an essential factor, trends are often forecast using time as the independent variable. Much of the "trend extrapolation" in futures research takes this form. Common methods of time-series forecasting being used today are the smoothing, decomposition, and autoregression/moving average methods. Smoothing methods are used to eliminate randomness from a data series to identify an underlying pattern, if one exists, but they make no attempt to identify individual components of the underlying pattern. Decomposition methods can be used to identify those components--typically, the trend, the cycle, and the seasonal factors--which are then predicted individually. The recombination of these predicted patterns is the final forecast of the series. Like smoothing methods, decomposition methods lack a fully developed theoretical basis, but they are being used today because of their simplicity and short-term accuracy. Autoregression is essentially the same as the classical multivariate regression, the only difference being that the independent (predictor) variables are simply the time-lagged values of the dependent (predicted) variable. Because time-lagged values tend to be highly correlated, coupling autoregression with the moving average method produces a very general class of time-series models called autoregression/moving average (ARMA) models.

All regression and time-series methods rest on the assumption that the historical data can, by themselves, be used to forecast the future of a series. In other words, they assume that the future of a trend is exclusively a function of its past. This assumption, however, will always prove false eventually because of the influence of forces not measured by the time series itself. That is to say, unprecedented sorts of events always occur and affect the series, which is precisely why the historical data are so irregular.

These difficulties have not deterred many traditional analysts and long-range forecasters from using such methods and thereby generating dubious advice for their sponsors. Within futures research, however, these techniques--when used well--are applied in a very distinctive way. The objective is not to foretell the future, which is obviously impossible, but to provide purely extrapolative base-line projections to use as a point of reference when obtaining projections of the same trends by more appropriate methods. What would the world look like if past and current forces for change were allowed to play themselves out? What if nothing novel ever happened again? The only value of these mathematical forecasting techniques in futures research is to provide answers to these remarkably speculative questions. But once they are answered, a reference will have been established for getting on with more serious forecasting.

For example, in a study by Boucher and Neufeld (I 98 1), a set of I I I trends was forecast 20 years hence both mathematically (using an ARMA technique) and judgmentally (using the Delphi technique). Analysis of the results showed that the average difference between the two sets of forecasts was over 15 percent. By the first forecasted year (which was less than a year from the date of the completion of the Delphi), the divergence already averaged more than 10 percent; by the 20th year, it had reached 20 percent. This result is interesting because even experienced managers usually accept mathematical forecasts uncritically. They like their apparent scientific objectivity, they have been trained in school to accept their plausibility, and acceptance has been reinforced by an endless stream of such projections from government, academia, and other organizations. Seeing judgmental and mathematical results side-by-side can thus be most instructive. Moreover, as some futures researchers believe, if the difference between such a pair of projections is 10 percent or more, it is probably worth examining in depth.

The Delphi technique
Given the limitations of personal forecasting (implicit or genius) and of mathematical projections, it is now common--and usually wise--to rely on systematic methods for using a group of persons to prepare the forecasts and assessments needed in strategic planning. Experience suggests, however, that at least five conditions must be present before the decision to use a group should be made: (1) No "known" or "right" answers exist or can be had (that is, acceptable forecasts do not exist or are not available); (2) equally reputable persons disagree about the nature of the problem, the relative importance of various issues, and the probable future; (3) the questions to be investigated cross disciplinary, political, or jurisdictional lines, and no one individual is considered competent enough to cope with so many subjects; (4) cross-fertilization of ideas seems worthwhile and possible; and (5) a credible method exists for defining group consensus and evaluating group performance.

The fifth condition is especially important--and often slighted. As a matter of fact, the emphasis one places on this consideration often determines the method of group forecasting one chooses. If, for example, the person seeking the forecasts will be content with an oral summary of the results (or perhaps a memo for the record), then a conventional face-to-face meeting of some sort may be the appropriate method. If, at the other extreme, it is known that the intended user will insist on having a detailed comprehensive forecast and that the persons whose views should be solicited would never speak openly or calmly to each other at a face-to-face meeting, then a different scheme for eliciting, integrating, and reporting the forecasts would surely be required.

Considerations like these were responsible in large part for the invention of what is no doubt the most famous and popular of all forecasting methods associated with futures research: the Delphi technique. Delphi was designed to obtain consensus forecasts from a group of "experts" on the assumption that many heads are indeed often better than one, an assumption supported by the argument that a group estimate is at least as reliable as that of a randomly chosen expert (Dalkey 1969). But Delphi was developed to deal especially with the situation in which risks were inherent in bringing these experts together for a face-to-face meeting--for example, possible reluctance of some participants to revise previously expressed judgments, possible domination of the meeting by a powerful individual or clique, possible bandwagon effects on some issues, and similar problems of group psychology. The Delphi method was intended to overcome or minimize such obstacles to effective collaborative forecasting by four simple procedural rules, the first of which is desirable, the last three of which are mandatory.

First, no participant is told the identity of the other members of the group, which is easily accomplished if, as is common, the forecasts are obtained by means of questionnaires or individual interviews. When the Delphi is conducted in a workshop setting--one of the more productive ways to proceed in many cases--this rule cannot be honored, of course.

Second, no single opinion, forecast, or other key input is attributed to the individual who provided it or to anvone else. Delphi questionnaires, interviews, and computer conferences all easily provide this protection. In the workshop setting, it is more difficult to ensure, but it can usually be obtained by using secret ballots or various electronic machines that permit anonymous voting with immediate display of the distribution of answers from the group as a whole.

Third, the results from the initial round of forecasting must be collated and summarized by an intermediary (the experimenter), who feeds these data back to all participants and invites each to rethink his or her original answers in light of the responses from the group as a whole. If, for example, the participants have individually estimated an event's probability by some future year, the intermediary might compute the mean or median response, the interquartile range or upper and lower envelopes of the estimates, the standard deviation, and so forth, and pass these data back to the panelists for their consideration in making a new estimate. If the panelists provided qualitative information as well--for example, reasons for estimating the probabilities as they did or judgments as to the consequences of the event if it were actually to occur--the role of the intermediary would be to edit these statements, eliminate the redundant ones, and arrange them in some reasonable order before returning them for the group's consideration.

Fourth, the process of eliciting judgments and estimates (deriving the group response, feeding it back, and asking for re-estimates in light of the results obtained so far) should be continued until either of two things happens: The consensus within the group is close enough for practical purposes, or the reasons why such a consensus cannot be achieved have been documented.

In sum, the defining characteristics of Delphi are anonymity of the estimates, controlled feedback, and iteration. The promise of Delphi was that if these characteristics were preserved, consensus within the panel would sharpen and the opinions or forecasts derived by the process would be closer to the "true" answer than forecasts derived by other judgmental approaches.

Thousands of Delphi studies of varying quality have been conducted throughout the world since 1964, when the first major report on the technique was published (Gordon and Helmer 1964). The subjects forecast have ranged from the future of absenteeism in the work force to the future of war and along the way have included topics as diverse as prospective educational technologies, the likely incidence of breast cancer, the future of the rubber industry, the design of an ideal telephone switchboard, and the future of Delphi itself. Some of these studies proved to be extremely helpful in strategic planning; a few virtually decided the future of the sponsoring organization. But most had little or no effect, apart from providing general background information or satisfying a momentary curiosity about this novel method of forecasting.

Part of the problem in many cases is that practitioners have had false hopes. The literature conveys the impression that Delphi is so powerful and simple that anyone can "run one" on any subject. What the literature often fails to mention is that no established conventions yet exist for any aspect of study design, execution, analysis, or reporting. Intermediaries, who are the key to useful and responsible results, are very much on their own. As novices they should examine studies by others, but because these studies are all different, it may be very difficult to find or recognize good models. Even with an excellent model in hand, the newcomer cannot fully appreciate what it means to use it. Only through practice can one discover the significance of four key facts about Delphi: (1) The amount of information and data garnered through the process can and will explode from round to round; (2) good questions are difficult to devise, and the better the design of the questions asked, the more likely it is that good participants will resign from the panel out of what has been called the BIF factor--boredom, irritation, and fatigue--because they will be asked to answer the same challenging questions again and again for each trend or event in the set they are forecasting; (3) the likelihood of such attrition within the panel means not that the questions should be cheapened but that large panels must be established so that each participant will have fewer questions to answer, which is very time consuming; (4) Delphi itself does not include procedures for synthesizing the entire set of specific forecasts and supporting arguments it produces, so that when the study is "completed," the work has usually just begun. And if, as one hopes, the intermediary and the panelists take the process and the questions seriously, the probability is high that the schedule will slip, the budget will be overrun, and so on and on.

Another reason that success with Delphi is hard to achieve is that, despite 20 years of serious applications, very little is known about how and why the consensus-building process in Delphi works or what it actually produces. No wide-ranging research on the fundamentals of the method has been done for more than a decade. According to Olaf Helmer, one of the inventors of Delphi, "Delphi still lacks a completely sound theoretical basis.... Delphi experience derives almost wholly either from studies carried out without proper experimental controls or from controlled experiments in which students are used as surrogate experts" (Linstone and Turoff 1975, p. v). The same is true today. The practical implication is that most of what is "known" about Delphi consists of rules of thumb based on the experience of individual practitioners.

For example, a goal of Delphi is to facilitate a sharpening of consensus forecasts from round to round of interrogation. And, in fact, there probably has yet to be a Delphi study in which the consensus among the participating experts did not actually grow closer on almost all of the estimates requested (as measured by, say, a decline in the size of the interquartile range of estimates). Yet the limited empirical evidence available on this phenomenon is replete with suggestions that increased consensus is produced only in slight part by the panelists' deliberations on the group feedback from the earlier round. The greater part of the shift seems to come from two other causes: (1) The panelists simply reread the questions and understood them better, and (2) the panelists are biased by the group's response in the preceding round of interrogation (that is, they allow themselves to drift toward the mean or median answer). The difficulty posed by this situation--which is far from atypical of the problems presented by Delphi--is that no way has yet been found to sort out the effects of these different influences on the final forecast. Accordingly, the investigator must be extremely careful when interpreting the results. Claims that Delphi is "working" are always suspect.

On the positive side, though again as a strictly practical, non-theoretical matter, Delphi appears to have a number of important advantages as a group evaluation or forecasting technique. It is not difficult to explain the essence of the method to potential participants or to one's superiors. It is quite likely that some types of forecasts could not be obtained from a group without the guarantee of anonymity and the opportunity for second thoughts in later rounds (certainly true when hostile stake holders are jointly evaluating the implications of policy actions that might affect them differently). Areas of agreement and disagreement within the panel can be readily identified, thanks to the straightforward presentation of data. Perhaps most important, every participant's opinion can be heard on the forecasts in every round, and every participant has the opportunity to comment on every qualitative argument or assessment. For this reason, it becomes much easier to determine the uncertainties that responsible persons have about the problem under study. If the panelists are chosen carefully, a full spectrum of hopes, fears, and other expectations can be defined.

Delphi has been used in many policy studies in higher education. In one case, it was used to determine priorities for a program in family studies (Young 1978). Nash (1978), after reviewing its use in a number of studies concerning educational goals and objectives, curriculum and campus planning, and effectiveness and cost-benefit measures, concluded that the Delphi is a convenient methodology appropriate for a non-research-oriented population. The technique has also been used in a number of planning studies (Judd 1972). For example, it was used as a tool for getting planning data to meet the needs of adult part-time students in North Carolina (Fendt 1978).