|Seminar on Futures Techniques|
-- a university of arizona course on methods and approaches for studying the future
You can select each entry individually for a particular technique, or you can scroll down the full listing. You can also print this page and get all 15 techniques on a single printing.
It is important to "ask
the right question" when dealing with any futures analysis. We are often
comfortable in current ways, or unable to ask relevant questions in a comfortable
setting. Backcasting allows this to be done by "leaping" to the future
through a vision or scenario, without regard to the method of getting
there. This way, preferred futures can be identified and evaluated independently
of the constraints of how to achieve the future situation. Then, after sufficient
discussion to determine where we would like to be, we can look at where we are
and determine the various methods/means of getting to the preferred future.
There are methods to just
look around and see what people are doing, as an indicator of change. The most
visible product is John Naisbitt and his book titled "Megatrends".
He looks at trends starting at the "grassroots" level by following newspaper stories in a variety of papers. He then compares the number of column inches on various topics in multiple locations, and spots trends early.
Similar studies have been done by looking at the frequency of appearance in the literature of certain events, in combination with qualifying words. With rapid and inexpensive searching on CD-ROM units, this is practical for anyone.
Other examples are books or journals that publish "current issues in management or environment", to see what leaders in the field are saying. One example from business might be "what management styles will be necessary for the 1990's? (where the underlying causes of their recommended management style suggest future trends - less administrative overhead, more baby boomers).
One caution on all these
relates to the possibly temporal nature of the data.
You normally use this for
"experts" rather than for the general population. If you don't know
something about the subject, the results are not worth much. You can make these
very quantitative, so you get ranges and averages. You can also use a special
technique, where you assign more "value" to someone else on a particular
question IF you know they are more expert than you are on that question. The
Delphi is useful to get things focused and on the table for discussion; it may
also dilute all views to the lowest common denominator - so the blind will follow
the blind, rather than develop new insights. It is good in its place, probably
overrated, more historic than current.
This is frequently used
technique, and can be used in many situations. You get a small group of people
together (7-9) and FOCUS on one general question for about 1.5 or so hours.
There is no structured agenda or set of specific questions. The facilitator
presents the topic (normally the group has been told of the topic in advance).
The group is normally diverse: selected from a variety of experiences and expertise
or training. Then the group discusses this general subject, with each person
learning a little from what the others say, and maybe making new comments they
have not previously thought of.
The advantage of this technique is that you keep digging at the issue until the group is satisfied you have really addressed it. Normally the 1.5 or so hours is sufficient, and if you use more than about 9 not everyone can talk (and if less than about 5 you get too little diverse opinion). These are used often by prime time television news - for example: How should we reduce services if we don't have enough revenue.
If you do 2-3 focus groups, with different groups of people on the same general topic, you will cover many views, gain from the "knowledgeable" people in different disciplines, and find a much improved understanding of the subject. It is a good technique. It was used in older days without a formal name - could be a group of folks going out for a beer one evening.
This technique is more focused,
and is generally done only by a narrow group of experts. Examples would be to
identify ranges of operational speed for commercial aircrafts in the next 20
years. In this example, it might involve looking at new engines, new metals
(or non-metallic composite materials because of strength and light weight),
or relatively minor changes to existing aircraft (e.g., more powerful engine
of same type).
In general, a new technology replaces an old one about the time the old one runs out of steam. Again the aircraft example, jet aircraft replaced the piston engine at the beginning of WWII. Supersonic aircraft replace some jets for long distance travel. This is an area where science fiction discussions can really help identify options for consideration. It is also a place where big errors can be made (because a narrow group - like social, political, economic). The SST (supersonic transport) in this country is a casualty of this lack of understanding. So, the technique is good for the area it is intended, but needs to have "additional" considerations to be a valid forecasting tool. The more broadly applied analysis is "technology assessment", which takes into consideration other factors).
This is where one person
interviews one person. It might last an hour or two, and requires GREAT skill.
The interviewer should be able to gain insights from a "casual" conversation
so the person being interviewed does not get too narrow in addressing a single
point (unless you want a lot of info about a specific thing). This is a bit
like the focus group, because you can keep asking questions until you get a
satisfactory response. These are expensive and hard to do well, so there are
not a lot of them. But, they are very good sources of information. and are especially
useful when certain you must understand the views of certain people (because
of their position or their expertise).
This method uses mathematical
relationships to explain a system, and then when it is understood, extrapolations
into the future can be made. The overriding consideration is that you "know
and understand" the system you are trying to model. In relatively simple
systems, or those that have been used a long time and have many revisions from
experience (like current economic forecasting), the relationships can be modeled
Two major components are 1) knowing the relationships relative to what event is connected to what other event(s); and 2) the relative magnitude of that relationship. In anything but a very simple model, the interactions and feedback loops (results of one step affect an earlier step) are VERY difficult to determine. This is especially true for models that predict more than a few years. For complex situations, it is nearly impossible to model accurately both the relationships and their magnitude when appropriate feedback loops are considered.
Models are useful for two reasons: 1) when you cannot make simple extrapolations or modify trends, you get better results with models even with their limitations; and 2) you can change the conditions and see what would happen under a variety of assumptions. The latter are "what if" options, and you learn a great deal about the subject and its future possibilities by determining which changes cause what type (and how big) of an effect. However, the model results are only as good as the model. Those that are used a lot on actual conditions so they can be tested with real data, and used for short term (few years) extrapolation are the best.
Scenario building is a relatively
recent technique, but widely used. A common technique is to 1) identify general,
broad, driving forces, which are applicable to essentially all scenarios, 2)
identify a variety of PLAUSIBLE trends within each driving force topic (trends
that vary depending on your assumptions so you get positive and negative perspectives),
and 3) combine the trends so you get a series of scenarios (for example, mostly
positive trends from all driving force topics would give a positive scenario).
The number of scenarios should be around 3-5, too fewer and you don't have a
range, and too many more and there are too many to understand. Generally, you
would find a positive (or optimistic), negative (or pessimistic), and neutral
(or middle of the road). Additional choices might be something involving an
unlikely event but one that would have a large impact.
Scenarios are good because you don't have to model complex situations, you can alter the combinations, and play "what if" games (e.g., change the assumption and see what happens). The major use is UNDERSTANDING the situation rather than trying to predict the future. As you work through various scenarios, you get an understanding of events and possible combinations. The disadvantage of scenarios is that it is hard to get agreement (if the questions are controversial) on what is the "right" scenario to include, and the problems of making it clear the scenarios are not firm predictions.
There are many ways to have
group involvement in making change or developing goals or other targeted activities.
The search conference approach began in the 1960s and improves on these techniques
but many of the features are found in other approaches. Two primary ways this
process differs from other group efforts is 1) rationalization and understanding
of conflicts, and 2) democratic process where the group is self directing. The
basic goal of a search conference is for a "community" of people to
plan their own future AND to generate a strategy and take the responsibility
for achieving it. The basic elements include 1) asking the right questions,
2) having the right mix of participants in the room at the same time, 3) building
implementation plans with consideration for barriers and driving forces of change.
Generally involves 20-35 people over a 2-3 day process. More details
This is not really a standard
futures technique. It is used more for "mind expanding" and allowing
you to get out of rigid patterns of thought, so you can look at the situation
from a fresh angle. Obviously it is mostly directed at "science" but
other behavioral patterns and non-science issues can be found in most science
fiction. By using imagination cleverly, through new terms to describe familiar
events or old terms to discuss new events (or new terms for new events), you
change you perspective. That is very valuable in looking at futures issues.
The single largest problem in futures understanding (my bias here) is to get
out of your old mindset and go "meta" to the issue (going meta allows
you to go beyond your normal self and take a look from entirely new perspectives).
Thus, this technique is good for developing ideas for some of the other more
specific techniques (especially something like scenario building).
This is good for "snapshots"
of what people's "perceptions" are at a given time. They are very
useful to get a feeling of what people think about issues. This indicates to
you if you should use an educational program of a certain type to change attitudes,
market your services to how people are already ready to receive them, or if
you should give up on what you were planning to do because the group you surveyed
is not ready for that.
If you want responses of this type, it is a very good technique. It is highly refined and can be done a number of ways. We have a group at the university that does telephone surveys, upon hearing the response immediately types the key answer into the computer, and the results are easily and rapidly summarized or cross tabulated with other answers to give more information. But, for serious ideas about future trends, new techniques, or other more detailed understandings of a subject, this is not the appropriate technique.
Trends can be easily extrapolated
into the future, or the basic trend modified (slightly) to reflect current and
anticipated changes. Any large system generally changes slowly due to simple
inertia, and thus on a short term basis this is a relatively non-risky procedure.
Where changes are relatively small, the variables known and generally unchanging,
and there is a long history, this is not a bad method.
It is especially weak when unanticipated events occur. The best recent example the energy industry and the initial oil embargo of 1973-74; this caused a perturbation in the forecasting (from straight line extrapolations) to something that is still hard to forecast today.
This is good for brainstorming as a group or individually. A concept or word is put on a large paper (or via several types of software - search Internet for "concept mapping" for software sources). Then you think of related words or concepts and link them to the initial word. Then you repeat this for each of the new words or concepts, until you have a large listing of ideas/concepts that all grew from asking questions about the central word (e.g., what does it mean, what does it impact, what impacts it, and so on). You will likely find a lot of duplication of ideas as this "interlinked web of words" gets large. You can then make a more concise map or summary of the key elements and how they are related. If you do this by software, the summarizing and especially the different ways to format output adds a new dimension to your understanding of the original concept.
Further information is available
from the Mind
Mapping FAQ and the Concept
Mapping home page.
This is a technique that summarizes the effects across several activities. Its primary value is that it allows for interaction across trends/events rather than treating them individually. Lets use 3 trends: 1) cost of electricity will increase substantially in next 10 years, 2) family income will remain constant in real dollars, and 3) nuclear waste from power plants will accumulate to a level where we need to dispose of it in new ways. The cross impact might look like:
----- Effect on These
Effect of Trend 1 Trend 2 Trend 3
Trend 1 x -- + (x indicates no match
Trend 2 - x - (effect as matches)
Trend 3 + -- x (self trend)
+ is slightly big effect, ++ is really big effect, - and -- are similar.
This is frequently qualitative but can be made quantitative and becomes complex. It is good for early analysis when you are better understanding the impacts of assumptions and interactions. It would then likely cause further analysis and study to better understand the implications. So, it is more of an internal technique to make a better product for presentation to others.
This technique is not well
known or widely used. It is mostly incorporated into brainstorming techniques.
However, it has been used in the educational area and was developed at the University
of Massachusetts. Basically, a small group of people produce an imaginative
list of future possibilities, then the group eliminates those that are highly
unlikely or too much like the present. Each remaining issue is developed in
detail (by other techniques) or group discussion). The initial item is then
developed into 3-6 subtopics, and then each subtopic is further divided into
3-6 more topics, until you are to a third of forth level. This technique is
finding use in the continuous improvement literature (i.e., Total Quality Management)
relating to methods of identifying problems and related events.
It is important to understand
the purpose for your evaluation of future conditions. Developing foresight capability
is not intended to create large employment opportunities for database developers.
To know what approach you want, answer the following questions: What will the
information be used for? Who will use the information and in what form? What
type of data will be collected and how will it be converted to useful information.
What alternatives exist to getting the same information (as opposed to specific
collection of data)?
Some examples of use for foresight studies: As input to developing alternative scenarios As background position papers or introductory sections in a variety of reports. As radar to indicate where the future might be heading. As a database to determine rapid "semi answers" to currently unknown questions (where data do not exist).