14
1. Introduction
though most of them are more concerned with descrip tion rather than prediction because of the fundamental limitations inherent in any effort to engage in predictive modeling.
Many of the decisions that society faces also requi re that its members are
effective in sharing their information and knowledge with each other- that
they communicate their assumptions abo ut system behavior and that they
ident ify the likely system responses und er alternative assumptions. Past experien ce, rules of thumb or feedback from opinion polls (e.g. Rich 1978)
are often used to arrive at a conclusion, and attempts are often made to justify and support a decision by soliciting expert advice (see e.g. Whitman
1985, Baumgartner and Midtturn 1987). Alternatively, consultants are hired
to find solutions to problems, or scientific studies are commissioned to
chart the likely behavior of the system of interest. Those studies are then
used as input into the decision making pro cess. In this case, it is typically
the expe rts who define the problem such that it can be addressed with their
problem-solving expertise. Once a problem is circumscribed, experts provide advice on how to address and solve it.
On e approach to societal decision making would be to identify a group
of experts and ask them for advice. This is typically done in management
decision making , whe n consultants are brought in to find solutions for
problems, and in policy decision making, when studies are commissioned
to chart the likely behavior of a socioeconomic, technological or environment al system. In either case , it is the experts who define the problem such
that it can be addressed with their problem-solving experti se. Once they
have nailed the problem down, they provide advice on how to address it.
If you ask different groups of experts, they may look at the problem differently and they may come up with different solutions. Once the experts
disagree, the que stion: What should I do? changes into the question : Which
advisor should I believe?This new que stion is often as difficult to answer as
the first one.
Of course, one could always add layers to this process, such as have an
advisory staff help with decision making or with the selection of advisors.
That approach would not solve your problem, but rather move it to a different level in the decision making hierarchy.
The advice on which experts base their judgment is typically derived
from models of the respective system. Scientists and consultants develop
databases and simulation tools to help managers and politicians make decisions. In some cases, these model s have not been developed from scratch
by the people who use them , but simply modified or combined to provide
an answer to a specific question . Noticing their reliance on models, you
may therefore be tempted to not ask expe rts for the answers that they generate with the ir model s, but instead ask them to give you their model s so
you can form your own opini ons.
Expert systems, simulation games, and learning laboratories are three examples of model environments produced by consultants and scientists to
1. Introduction
though most of them are more concerned with descrip tion rather than prediction because of the fundamental limitations inherent in any effort to engage in predictive modeling.
Many of the decisions that society faces also requi re that its members are
effective in sharing their information and knowledge with each other- that
they communicate their assumptions abo ut system behavior and that they
ident ify the likely system responses und er alternative assumptions. Past experien ce, rules of thumb or feedback from opinion polls (e.g. Rich 1978)
are often used to arrive at a conclusion, and attempts are often made to justify and support a decision by soliciting expert advice (see e.g. Whitman
1985, Baumgartner and Midtturn 1987). Alternatively, consultants are hired
to find solutions to problems, or scientific studies are commissioned to
chart the likely behavior of the system of interest. Those studies are then
used as input into the decision making pro cess. In this case, it is typically
the expe rts who define the problem such that it can be addressed with their
problem-solving expertise. Once a problem is circumscribed, experts provide advice on how to address and solve it.
On e approach to societal decision making would be to identify a group
of experts and ask them for advice. This is typically done in management
decision making , whe n consultants are brought in to find solutions for
problems, and in policy decision making, when studies are commissioned
to chart the likely behavior of a socioeconomic, technological or environment al system. In either case , it is the experts who define the problem such
that it can be addressed with their problem-solving experti se. Once they
have nailed the problem down, they provide advice on how to address it.
If you ask different groups of experts, they may look at the problem differently and they may come up with different solutions. Once the experts
disagree, the que stion: What should I do? changes into the question : Which
advisor should I believe?This new que stion is often as difficult to answer as
the first one.
Of course, one could always add layers to this process, such as have an
advisory staff help with decision making or with the selection of advisors.
That approach would not solve your problem, but rather move it to a different level in the decision making hierarchy.
The advice on which experts base their judgment is typically derived
from models of the respective system. Scientists and consultants develop
databases and simulation tools to help managers and politicians make decisions. In some cases, these model s have not been developed from scratch
by the people who use them , but simply modified or combined to provide
an answer to a specific question . Noticing their reliance on models, you
may therefore be tempted to not ask expe rts for the answers that they generate with the ir model s, but instead ask them to give you their model s so
you can form your own opini ons.
Expert systems, simulation games, and learning laboratories are three examples of model environments produced by consultants and scientists to