Summary Thinking and Decision making
Chapter One: Introduction and Overview
Rational behaviour implies that people behave according a particular normative model.
However, in many situation, people do not act rationally. Economists have argued that as
people gain experience of a particular domain they learn accordingly and behave in a more
rational way. There is some evidence to that but also a lot of evidence showing the opposite
effect. However, violations of rationality might not always be considered as irrational.
Evolutionary related behaviour may nonetheless not predict behaviour in certain
specific instances. Illusions occurring in the environment makes us perceive object differently
than they really are, which isn‟t necessarily a bad thing. It helps us navigate through our
environments. Still, both the intellect and visual system can pose problems and the same goes
for judgment and decision making. People tend to see sequences that are non-existent and
people often attend to immediate consequences instead of future ones.
It seems that people often don‟t behave in their own best interest, so it is tempting to
think of them as irrational. On the other hand, the success of the species as a whole suggests
otherwise. Some authors state we shouldn‟t focus on rationality at al. Instead, we should take
beliefs, preferences and constraints into account. This is also knows as bounded rationality.
Simon (1959) believed the mind has evolved short-cut strategies that deliver reasonable
decisions, better known as heuristics. Both the negative and the positive effects have been
thoroughly researched. In an interactive point of view, the heuristics often lead to positive
outcomes precisely because the heuristics are the product of environmental circumstances. A
similar approach is the one of the adaptive decision maker. People use particular strategies
depending on a trade-off between effort and the need for accuracy.
We cannot however, forget the influence of the unconscious mind. Both the conscious
and unconscious have some advantages and, as will be discussed in chapter 15, many
researches propose a dual system theory.
, Chapter 2: The nature of analysis of judgment
Social Judgment Theory:
Judgments depend on the recognition of certain cues, and knowing how much influence they
have on a particular situation. This is implemented in the Social Judgment Theory. By using
statistical analysis to a series of these judgments it is possible to describe the impact that
different cues have on a particular type of judgment. When the outcomes are known in a
certain case we speak of a predictive model. However, when using a multiple regression
analysis, i.e. the values are unknown and calculated by co-occurrence, we speak of a
Bootstrap model. The lens model is nice way to depict this mechanism. The to-be-judged
criterion is predicted by some environmental cues, carrying some predictive value or
ecological validity. The processing and consideration of all these cues influence the judgment
that will be made.
Unaided human judgment is often referred to as clinical prediction. Clinical and
actuarial (statistical) predictions have often been compared, which showed actuarial
predictions are superior. Even when people have more information available, actuarial
predictions are a lot more accurate. This doesn‟t mean that humans are unimportant in the
judgment process, because the correct information still has to be selected and coded. However,
people do not always apply the cues in a correct manner. These errors will all be discussed in
the following chapters. People can however detect changes in the environment whereas
actuarial models cannot. Composite models of peoples predictions were more accurate than
solely actuarial or clinical predictions alone.
Bootstrap models are actually improper linear models, because they do not optimally
weight the cues. In smaller samples or without measurable criterion values, multiple
regressions are no longer reliable. Random and equal weighted models appear to be better
than the bootstrap model.
This all indicates that if possible, humans must be replaced by models. However, this
has never had the impact it should have. Lots of professionals are very reluctant because it
questions their expertise. Besides that people have a selective memory for past cases and think
that every situation is unique so that statistics do not apply. Unfortunately does this extra
amount of information no particular good in for accuracy.
Chapter One: Introduction and Overview
Rational behaviour implies that people behave according a particular normative model.
However, in many situation, people do not act rationally. Economists have argued that as
people gain experience of a particular domain they learn accordingly and behave in a more
rational way. There is some evidence to that but also a lot of evidence showing the opposite
effect. However, violations of rationality might not always be considered as irrational.
Evolutionary related behaviour may nonetheless not predict behaviour in certain
specific instances. Illusions occurring in the environment makes us perceive object differently
than they really are, which isn‟t necessarily a bad thing. It helps us navigate through our
environments. Still, both the intellect and visual system can pose problems and the same goes
for judgment and decision making. People tend to see sequences that are non-existent and
people often attend to immediate consequences instead of future ones.
It seems that people often don‟t behave in their own best interest, so it is tempting to
think of them as irrational. On the other hand, the success of the species as a whole suggests
otherwise. Some authors state we shouldn‟t focus on rationality at al. Instead, we should take
beliefs, preferences and constraints into account. This is also knows as bounded rationality.
Simon (1959) believed the mind has evolved short-cut strategies that deliver reasonable
decisions, better known as heuristics. Both the negative and the positive effects have been
thoroughly researched. In an interactive point of view, the heuristics often lead to positive
outcomes precisely because the heuristics are the product of environmental circumstances. A
similar approach is the one of the adaptive decision maker. People use particular strategies
depending on a trade-off between effort and the need for accuracy.
We cannot however, forget the influence of the unconscious mind. Both the conscious
and unconscious have some advantages and, as will be discussed in chapter 15, many
researches propose a dual system theory.
, Chapter 2: The nature of analysis of judgment
Social Judgment Theory:
Judgments depend on the recognition of certain cues, and knowing how much influence they
have on a particular situation. This is implemented in the Social Judgment Theory. By using
statistical analysis to a series of these judgments it is possible to describe the impact that
different cues have on a particular type of judgment. When the outcomes are known in a
certain case we speak of a predictive model. However, when using a multiple regression
analysis, i.e. the values are unknown and calculated by co-occurrence, we speak of a
Bootstrap model. The lens model is nice way to depict this mechanism. The to-be-judged
criterion is predicted by some environmental cues, carrying some predictive value or
ecological validity. The processing and consideration of all these cues influence the judgment
that will be made.
Unaided human judgment is often referred to as clinical prediction. Clinical and
actuarial (statistical) predictions have often been compared, which showed actuarial
predictions are superior. Even when people have more information available, actuarial
predictions are a lot more accurate. This doesn‟t mean that humans are unimportant in the
judgment process, because the correct information still has to be selected and coded. However,
people do not always apply the cues in a correct manner. These errors will all be discussed in
the following chapters. People can however detect changes in the environment whereas
actuarial models cannot. Composite models of peoples predictions were more accurate than
solely actuarial or clinical predictions alone.
Bootstrap models are actually improper linear models, because they do not optimally
weight the cues. In smaller samples or without measurable criterion values, multiple
regressions are no longer reliable. Random and equal weighted models appear to be better
than the bootstrap model.
This all indicates that if possible, humans must be replaced by models. However, this
has never had the impact it should have. Lots of professionals are very reluctant because it
questions their expertise. Besides that people have a selective memory for past cases and think
that every situation is unique so that statistics do not apply. Unfortunately does this extra
amount of information no particular good in for accuracy.
