Longer version

Week 5 : Wednesday 1st November
Lecturer: Jennifer Harper
Topic: Foresight Methodologies and
Clustering of Methods by Task
Categories of Methods
Foresight and Forecasting Methodologies
Methods by Phases
Exploratory vs Normative
Other methods
Foresight methods and processes draw on futures,
planning and networking approaches.
EU-US approaches differ : EU more fragmented and
diverse as are the methods used
Identify the strengths & weaknesses of the different
approaches and address the concerns
Ensure appropriate use of methods/ Combined use of
methods recommended. Choice and combination of
methods have important implications in terms of process,
results and policy impacts
Ethical concerns: exploratory/creative vs. normative,
In-built paradox: predicted futures vs invented/created
futures; adaptive change and the self-fulfilling prophesy
Clustering of Methods by Task
1. Methods to manage foresight
ROAME (Rationale, Objectives, Activities,
Monitoring and Evaluation)
 Set systems to document the process + outputs
- The structures – steering committee
2. Pre-Foresight Methods
Identifying stakeholders – co-nomination
Environmental Scanning
Adapted from Miles 2002
Clustering by Task
3. Foresight Generation
Genius forecasting, Expert panels, Brainstorming,
Relevance trees, cross-impact analysis, delphi,
scenarios, trend extrapolation…
Key technologies
4. Foresight action and renewal
Action Plans, Demonstrators, Panel Embedding,
Publications, Outreach
Adapted from Miles 2002
Categories by Criteria
Quantitative methods (use of
statistics and other data) to
elaborate future trends and impacts
− Trend extrapolation
− Simulation modelling
− Cross impact analysis
− System dynamics
Qualitative methods (drawing on
expert knowledge) to develop long
term strategies
− Delphi method
− Experts panels
− Brainstorming
− Mindmapping
− Scenario analysis workshops
− SWOT analysis
Methods to identify key points of
action to determine planning
− Critical/ key technologies
− Relevance trees
− Morphological analysis
Source: FOREN Guide
Categories of Methods
Quantitative methods provide numerical
representation of future developments, e.g.
forecasts and modelling. They give a great deal
of weight, an ability to examine rates and scales
of changes but they limit the comprehension of
social and political variables and are not always
Qualitative methods vary from creative thinking
to more systematic techniques. Used when data
are hard to collect or not available.
Foresight can never be completely
dominated by quantitative methods: the mix
depends on access to relevant expertise and
the nature of the issues.
Source: Foren Guide
Categories of Methods
Exploratory methods (“outward bound”) start
from the present and move forward to examine
kinds of alternative future developments it can
lead to. Look at implications of possible
developments that lie outside familiar trends:
What if methods
Normative methods (“inward bound”) start with
the creation of a preliminary view of possible
(preferred) futures, then an analysis of how
these futures may grow out of the present asking what trends and events would take us
there or can be avoided. How methods.
Source: Foren Guide
Foresight and forecasting methodologies
3 processes of precedent,
analogy and extrapolation
that gather material by
looking back/backcasting
based on the assumption
of a stable world where
tomorrow will be different
from today just as today
was different from
yesterday, but that the
same things always
change in the same way
at the same rate,
following some
immutable law, or
demographic change.
More qualitative methods
based on eliciting expert
Shift from refined
sectoral forecasts “better to sweep wide
and large to glean the
macro-trends than to
forge highly
sophisticated tools for
segments of realities,
which generate
quantitatively precise
forecasts that are
generally wrong because
they skip over
Foresight and forecasting methodologies
Forecasting main tools:
Traditional methods of
econometric model
based on mechanics
and transposed to
microeconomics and
then macroeconomics.
Simulation modelling
Cross impact analysis
System dynamics
Foresight Main tools:
Delphi method
Expert panels
Scenario analysis
and building
Critical technologies
Relevance trees
Methods by Phases
Define the problem:specific challenges and needs, e.g. a
longer-term innovation strategy), set the time horizon.
Methods: Panels, SWOT or brain-storming,
Scope -possible focus on specific sectors, bottom-up vs
top-down, extent and structure of consultation
(experts/non-experts), resources, choosing appropriate
Identify the key variables analyzing the relations between
variables, by placing them in matrix to identify drivers.
Methods: panels/ interviews/desk-based research/
specialized studies/ cross-impact analysis: influence and
dependency graph.
Gathering data on trends and drafting hypotheses.
Methods: trend analysis, simulation modeling, weak
Exploring possible future through scenario-building
Methods: Exploratory or Normative approaches, panels 11
Scoping for a regional exercise
What are the possible futures of the region in
terms of strengths and weaknesses?
What sectors, technologies and industries
should be promoted to enhance
What are the likely prospects for regionally
important markets in general?
What are the likely changes affecting the market
important to a specific regionally significant
What are the likely development trends in
central social issues?
Source: FOR-RIS Report 2004
Quantitative Methods
Forecasting (time series analysis)
Trend Extrapolation (long-term vs
Cross Impact Analysis
Simulation Modeling
System Dynamics
Time series analysis
inherent laws in economic activity (quantified through past
these laws steer the development of economically relevant
variables onto certain ‘predestined’ courses and enable
visualization of future development.
Breakdown of a time series into its components:
typical seasonal but still short-term fluctuations,
business cycle
irregular influences
Short-term vs long-term (UK’s growth rate round 2%
since 1860)
Heavy trends vs seed events (weak signals 9/11)
Source: Hans Georg Graf
Types of trends
Linear Trend increases/decreases over time at constant absolute
rates of change.
Exponential Trend increases/decreases at constant percentage
rates of growth
Parabolic trend has the tendency of initially increasing or
decreasing only slightly and then more and more so with time.
Logistic Curves, Growth Curves are usually S-shaped, i.e.,
moving from an initial phase into a dynamic phase in order to
then enter into a new (stationary) equilibrium phase again in the
Life Cycle Curves follow at first a growth curve trend. In a
subsequent period the curve takes a downturn again. forecasts
are based on extrapolations of past trends.
Best fit - based on analysed series of numbers, ‘typical’ trend or
‘trend of the best fit’ is extrapolated into the future by assuming
that the relationships so expressed remain the same and
continue to exist in the same form in the future.
Source: Hans Georg Graf
Trend Extrapolation
The basis of extrapolation is to locate a trend that is
apparent over time, and project forward data re the
rates of change and the extent of change achieved.
In the shorter-term forecasts this is often a matter of
extending a linear or exponential curve – e.g.
economic growth, power or diffusion of a
In the longer-term, limits to growth will often be
encountered – there may be a limit to the size of the
population to whom a technology or cultural practice
can diffuse for example, and various other types of
trend curve may be fitted to the data (e.g. S-curve).
Adapted from Hans Georg Graf
Concerns with trend analysis
Superficial dichotomy of past as knowable unquestionable facts
and future as arbitrary opinion. Past is open to re-interpretation.
How reliable is trend data? Limited historical data or trends
guessed at.
How to find link between variables? GDP/employment by
Time sequences: the fertility rate in Fr over past two centuries
offers diametrically opposed trends depending on time sequence
Persistence of underlying driving forces? Will counter-trends
come into play, e.g. as resistance grows to a particular cultural
development, or competitors learn to challenge the power behind
a trend?
Adapted from Hans Georg Graf
Concerns with trend analysis
At what point will ceilings or turning points be
reached? A key challenge is estimating the ceiling in
diffusion of technologies or practices, e.g. level at
which the population is saturated with this new
product. Infant mortality as indicator of progress in
life expectancy (ceiling).
Is the quantitative trend masking qualitative change?
E.g. new technologies diffuse, but also change, e.g.
computers, cultural practices, diseases: implications
of later adoption of the new product or practice are
liable to be very different from those of the early
adoption. The skills required, the cultural meaning
are all liable to have changed.
Adapted from: Hans Georg Graf
Perceptions and misperceptions
Our view of reality depends on:
Our lens/blinkers (partial/blurred vision)
Mindsets, mental constructs, emotion, bias
Our means of observation and knowledge
sources available to us
The means of measurement (GDP)
The weight of the theories we use
The influence of ideologies /schools of thought
Source:Blueprints Report
Source: Hans Georg Graf
Cross Impact Analysis
An expert-based and quantitative method used in
preparation of scenarios, for establishing the
drivers and likely events.
Experts rate the likelihood of various events
occurring or not occurring.
The matrix of possibilities that arises from this
can be subject to mathematical analysis to
generate a list of scenarios, each with an
aggregate probability of occurrence assigned to
it. The cross-impact method forces attention to
chains of causality: ‘x’ affects ‘y’; ‘y’ affects ‘z’
Source: FOREN Guide
Utility of Cross-Impact Analysis
According to William Alexander, Cross-Impact
Analysis assesses the validity and accuracy of
a vision through a systemic approach:
Breaking vision down into specific “themes” &
Revealing interdependencies among themes and
Determining the probabilities of the relative
impact interrelationships may have on each
Calculating an overall “matrix index” that
provides a measure of the projected strength
and influences of the event interrelationships
and their likelihood of bringing about the
strategic future
Cross Impact Matrix
How does
this 
Top executives of highprofile corporations
employ deceptive
practices to rob
shareholders and other
The forces of economic
globalization make
national borders
permeable to commerce
and enable free trade.
Americans tend to value
Permeable borders and
free trade create more
opportunities for
deceptive business
government weakens
federal regulatory power
and creates a divide and
conquer phenomenon
(corporations can seek
the weakest local
Top executives of highprofile corporations
employ deceptive
practices to rob
shareholders and other
The forces of economic
globalization make
national borders
permeable to commerce
and enable free trade.
Corporate malfeasance
could make lawmakers
wary and incline them to
impose barriers to
international commerce.
Americans tend to value
There may be more
demand for centralized
federal control of
business practices.
American support for
government will
promote laissez-faire
trade while accelerating
trends toward
Free trade is compatible
with decentralized
government so freetrade advocates will
press for more
Concerns with
Cross Impact Analysis
The definition of events to be included in the study
can be a tiring and tedious process but is a crucial
step in a cross-impact analysis.
Any influences not included in the event set, will be
completely excluded from the study. However, the
inclusion of events that are not pertinent can
complicate the analysis unnecessarily.
Limited use as only suitable for small sets of
variables and very time-consuming for experts.
Limited independent analysis of its utility.
Source: FOREN Guide
Simulation Modeling
Modelling : both computer simulations and role-playing.
Computer simulations tend to be complex, representing a
system in terms of a range of variables and their interrelationships.
Large, all purpose vs simple issue-focused models
Equilibrium vs dynamic process-based models
Modelling has been most developed around relatively
easily quantifiable issues, e.g. economic growth,
employment, energy use, and demographics.
In recent years important modelling efforts have been
undertaken in examining climate change and
environmental impacts.
Modelling social, political and cultural change is much
more contentious.
Source: FOREN Guide
Concerns with Modeling
Who validates the data and relationships assumed?
dependent experts able to assess the quality of the
modelling effort?
Can key assumptions be debated, even by non-experts?
Is the model over complex?
Can it be made less costly to run, in terms of computer and
labour resources?
Is the model able to cope with structural or qualitative
changes on the horizon? Does it address future
developments within its framework?
Are the results dictated by the assumptions made?
Does the model assume that an equilibrium state is to be
reached? If so, is this remotely realistic ?
Source: FOREN Guide
Qualitative Methods
Expert panels
Scenario analysis and building
Critical technologies
Relevance trees
Delphi technique is a method used to reach
consensus in groups. Its advantage is that it
avoids group situations which are often prone to
control by individuals that have the best group
interaction and persuasion skills.
This method overcomes problems of face-toface confrontation in the group, by keeping the
responses and respondents anonymous.
Delphi method produces a rapid narrowing of
opinions and provides more accurate forecasts
than group discussions.
The Delphi Project
The Delphi Method Produces a Consensus-Based Response
Among the first group processes, developed by
RAND researchers in 50s to amalgamate expertise
from a wide range of knowledge areas and divergent
views to achieve eventual consensus.
Delphi process is iterative - in successive rounds, a
group of experts is asked to supply responses to a
list of questions. At the conclusion of every round,
the participants view each other’s answers and may
then change their views in light of what others
believe. The answers are presented anonymously to
eliminate the possibility that undue weight will be
placed on the responses of persons who hold
particularly high status within the group.
Chronology of delphi activity
Source: http://www.unido.org/file-storage/download/?file_id=16959
The Delphi Process
Organising a Delphi
Deciding if it is appropriate method
Defining the objectives (a process that may require use of other
methods like creativity procedures or scenarios or future
Setting up of expert panels (virtual?)
Shift to use of On-line delphi, e.g. Euforia and Fistera projects
Formulating the statements
Structuring the fields and defining the topics (mix of desk-based
research and creativity methods, brain-storming)
Setting of criteria: estimated time of realisation (normal time
horizon of delphi 30 years), level of expertise, which country is
leader in this field, which policy measures should be taken? Also
open questions.
In designing the questionnaire, need to consider how to give
feedback to experts in second round and also final presentation of
data to clients.
Constraints: complex logistics, time and resource intensive
Source: FOREN Guide
Delphi Results:
Measures and topic clusters
Source: Kuhls, 2001
Concerns about Delphi
Dependency on opinion of selected experts
Level of expertise difficult to assess
Poor execution leads to poor results
Results may be manipulated to produce desired
results in the second round
Good for single issues but poor on considering
issues within a holistic context (this is where
cross-impact analysis is important) – less useful
on complex forecasts based on multiple factors
Framing of statements subject to cultural and
social bias
Source: FOREN Guide
Step1 : a period of freethinking for articulation of ideas.
Step 2: collation of ideas and spin-off ideas. Not
normally subject to critique or further analysis
once a sufficient number generated. This
reduces inhibitions to consider “wild” ideas.
Step 3: ideas are grouped and prioritised and taken as
the basis for more analytic discussion.
Widely used technique for expert committees and
consultation management groups. Several computer
systems are now available as alternatives to the
traditional use of whiteboards and flipcharts to
record and display the ideas.
Source: FOREN Guide
• Technique applied to brainstorming and other
group discussion methods – useful for mapping
relationships between a large number of factors
• quick charting of group’s ideas in logical
groupings, even when ideas emerge nonsequentially
• allows efficient brainstorming for ideas, and at
the same time create a skeletal framework for
later categorisation of the information
• Mindmapping can be used in planning,
identifying customer groups. Mindmapping
works well when issues have many components
and subcomponents. This technique is a nonSource:
of outlining information.
A scenario is 'the description of a possible future and
the corresponding path to it' (Godet)
Godet presents 5 basic attitudes to the future:
1. the “passive” ostrich (go with the flow)
2. the “reactive” firefighter (damage reduction)
3. the “preactive” insurer(accident prevention)
4. the “proactive” innovative conspirator
5. the “anticipative” actor (blend of 2,3 &4)
Theser scenarios are normative, describing preferred
Shift from trend-based scenarios to scenarios
depicting deep breaks or even breakdown in current
Exploratory (outward)
Moving from present to the future
What next? What if?
Normative (inward)
Moving from future back to the present
Where to? How to?
Building scenarios
Concerns in Scenario Analysis
Superior to many other methods where number of
factors to be considered and the degree of
uncertainty are high
Stimulates strategic thinking, communication and
organisational agility
Difficult to draw up credible and useful scenarios
Users might find it difficult to deal with multiple
images of plausible futures
May be mistaken for forecasts
Strengths, Weaknesses, Opportunities, Threats
Used in Scenario workshops and Delphi
Focus on the current situation
Prepared by an expert team based on evidence
drawn from interviews, statistical or benchmarking
Assessment can be presented in 2x2 matrix
Main internal and external factors influencing strategies (or
possible futures)in positive or negative ways.
S: internal positive factors O: external positive factors
W: internal negative factors T: external negative factors
Critical Technologies
What are critical technologies?
Related to national self-sufficiency
generic and pre-competitive
High potential of various product applications.
How to identify list of critical technologies?
Apply criteria to measure importance of certain technologies –
clarity and consistency!!
Interviews, benchmarking, brainstorming, panels,
Indicate points of policy intervention
Setting the Criteria:
Be clear on what constitutes ‘critical”
capabilities vs needs, technology push vs supply or needs
Adapted from Foren Guide
Set of criteria used to assess importance & feasibility. Each criteria to
be assigned a mark from the scale 1-5 (1 = low; 5 = extremely high)
Social Importance
Importance for GDP
Importance for human
Impact on material
Importance for Export
Importance for the
safety of society
Impact on energy
Impact on productivity
Impact/influence on
the quality of life
Market size
Influence on the
creation of job opport.
Potential of replacing
non-renewable energy
by renewable ones
Strategic importance
for CR internationally
Natural and productive
space saving effect
Effect on transport
Research & technological opportunities
Ability of the R direction to produce new technologies
Probability of ‘breakthrough discoveries’
Probability of creation of new application possibilities related to
the R direction
Possibility of combining the R direction with other R directions
Possibility of applying the results of the R direction in various
Probability of a synergic effect with other research directions
Probability of involvement in international
Importance for meeting untackled needs of the society
Application potential
(absorption potential of the
application sector)
Research and technology
potential (production potential
of R&D)
Competitiveness of the application Current state of the art of the
sector (s)
research field
Support in admin/state policy and
Probability of a positive
development of the research field
Availability of results in the world
Level of necessary R &
Demand of the application sector
Financial reqmnts of the research
Probability of financing from various
Level of education in related fields
Current quality of human resources
Czech foresight https://www.unido.org/file-storage/download/?file_id=16961
Influence on the creation and
growth potential of SMEs
Relevance Trees and
Morphological Analysis
• Best-known normative forecasting methods
• Used to identify the circumstances, capabilities,
actions and knowledge to achieve future objectives
• Relevance trees sub-divide topic into sub-topics in a
tree-like structure, to map various aspects of a
system, a problem or solutions.
• Morphological Analysis maps all possible solutions
to a problem to determine alternative futures, e.g.
product development or building scenarios.
• These approaches are time-consuming as they
require in-depth analysis and multi-disciplinary
Adapted from Miles 2002
Horizon Scanning
Horizon scanning is defined as the systematic
examination of potential threats, opportunities and likely
future developments, including (but not restricted to)
those at the margins of current thinking and planning.
Horizon scanning may explore novel and unexpected
issues as well as persistent problems or trends.
UK Chief Scientific Adviser’s Committee, September 2004
[…] excellent horizon-scanning of current science and
technology, looking at opportunities and threats at least
five to ten years ahead, and often considerably beyond
that, is essential to the effective governance and
direction of Government policy, publicly funded research
and many of the activities of the private sector, and to the
interactions between them
UK Science & innovation investment framework 20042014
Environmental scanning
Environment refers not just to natural environment –
business, political, technology.
Approaches for identifying important future
developments in the environment of organisations:
Why: trend-spotting: in fashion, tastes, key players
How: issue surveys based on polling experts.
 “genius forecasting” experts write about topics
that they believe will be important for the future.
 systematic analysis of documentary sources, e.g.
press coverage or web searches means to chart
emergence of key themes.
 bibliometrics and patent analysis
Concern: Lock-in to particular sources/methods
Criteria for selecting methods
Organisational context
Nature of issue(s) under consideration
Quantitative/Qualitative data requirements
Time horizon
Methodological competence
Process/Product balance
Suitability for combining with other methods
Suitability for visualising the results
Source: Scapolo (EFA) 2003
For discussion
The ethics of forecasting and foresight
depends on the choice and combination of
methods used.
The balance between predicted futures
and created futures is a reflection of sociocultural context and local competencies in
use of futures methods.
Topics for Individual assignments
Assignments are to be handed in by end of January.
Topics that can be addressed apart from set questions at
the end of lecture:
Evaluation or impact assessment of a major foresight
Effective engagement /outreach to policy-makers and
Emerging themes in foresight like ethical issues,
systemic and/or adaptive foresight, Higher Education or
university foresight, regional foresight,
Emerging tools: disruption scenarios, success scenarios,
on-line approaches and horizon scanning.
FOREN GUIDE http://forera.jrc.es/documents/eur20128en.pdf
Hans Georg Graf