p.4 A decision maker needs only four types of information to construct a
decision tree:
1. What are my possible courses of action? (Alternatives)
2. What are the events that might follow from these actions? (Outcomes)
3. What is the likelihood of each event?
4. What is the value of each event to me?
p.6 One difficulty often encountered in constructing a decision tree is that likelihoods and utilities are
often not easy to assess... Because the helpfulness of a decision tree is based largely on the accuracy of the likelihoods
and utilities used, every effort should be made to obtain good estimates.
p.7 In other words, judgments are made from tangible data, which serve as cues
to intangible events and circumstances.
p.10 The simplest and best way to discover the cues, differential weights, function forms, and consistency
of a person's judgment process is to use a computer to present a number of cases to the person making the judgment. After
the judgments have been made, a computer program can readily decompose the judgment process into weights, function forms,
and consistency.
p.115-116 We define an objective as a statement of something that one wants to achieve... We distinguish
between fundamental objectives and means objectives. Fundamental objectives are concerned with ends rather
than means. Ends are the fundamental objects of value that stakeholders care about in a specific decision context, while means
objectives are methods to achieve ends... Strategic objectives should have the following characteristics. They should
be relevant to a wide range of decision contexts, to a long time period, and to many levels in an organization... Strategic
objectives should be deliberately general to enhance flexibility, which comes at the cost of detail.
p.121 For each of the lowest level objectives... we need to specify an attribute to measure the
degree to which the objective is achieved. Specifying attributes is the most effective way to make the objectives
hierarchy useful for subsequent analysis of strategic decisions... With attributes, the objectives can be used as a framework
to evaluate alternatives.
p.305 A fundamental and critical component of expert judgment is the ability to appropriately use
available information which varies in its relevance.
p.324-325 The expert must identify information or cues from the multidimensional stimulus he encounters.
These cues are diagnostic (i.e., contain information) about the final decision or judgment... While each
cue is related to the final categorization, it also serves the function of leading the expert to other cues
that will also be of importance in the determination of the final categorization... While cues are being identified
and clustered together, the expert is measuring the amount of the cue... After cues are identified,
measured, and clustered, the important cognitive work of weighting and combining to form a global evaluation
follows.
p.349 To make a sensible decision one needs to make two sorts of predictions: What will happen if
I do X? And will I like it?
p.390 Decision [making] involves predicting our future "states of mind."
p.391 evaluations and weights [in a linear model of decision making] that are "reasonable" provide outcomes
very close to those based on optimal ones.
p.395 In any field requiring judgmental forecasts, the performance of professional forecasters depends
jointly on (1) the environment about which forecasts are made, (2) the information system that brings data about the environment
to the forecaster, and (3) the cognitive system of the forecaster.
p.399 Since the forecast event is not fully determined by the cues, it cannot be predicted with
certainty. Similarly, because some inconsistency is pervasive in human judgment, forecasts also are not perfectly
related to the cues. Therefore, the relations on both sides of the lens model are probabilistic; that is, there is
an element of uncertainty in the relation between the cues and both the observed event and the forecast.
p.406 Forecasting skill may be degraded if the information system that brings data to the forecaster
does not accurately represent actual conditions.
p.407 An obvious way to improve the environment/ forecaster match is to have the forecaster learn
more about the nature of the environmental system... Increasing the forecaster's experience with the task... is another
obvious way to improve
p.409 Reliability of information acquisition is the extent to which the forecaster can reliably
interpret the objective cues. The evidence suggests that unreliability of information acquisition is pervasive.
p.609 When people are asked to generate an estimate, they frequently anchor on an obvious or convenient
number (e.g., the mean, the mode) and then adjust upward or downward if there is a reason to believe that the correct number
should be moved in either direction. In many situations, that strategy works well
