Judgement involves deciding on the likelihood of various events
using whatever information is available. What matters is accuracy.
Support theory
• A possible future event seems more likely when it is supported by
extra information. This is because people fail to remember all the
relevant information when it is not explicitly mentioned.
• The observation that we tend to overestimate the likelihood of
information that can easily be retrieved from memory and underestimate
the likelihood of information that is hard to retrieve is
called the availability heuristic.
Base-rate information
• People often make little or no use of base-rate information when
making judgements. This is particularly true when they are getting
information they have no personal experience with and when
likelihoods are expressed as probabilities.
• The observation that we pay too much attention to individuating
information at the expense of base-rate information is one example
of the use of the representativeness heuristic, the tendency we
have to decide that an object belongs to a given category simply
because it appears typical or representative of that category.
• Ignoring base-rate information can be overcome by deliberate
thinking, which is more likely when we question a judgement
(motivated reasoning) or when the problem formulation emphasises
the importance of differences in base rate. Judgement also
tends to be better when the problems are formulated in frequencies
rather than probabilities or when visual aids are used.
Why do we keep using heuristics?
• Laboratory studies have shown situations in which people are
easily misguided by their heuristics. In daily life, however, heuristics
are helpful most of the time.
• Humans seem to have a toolbox of fast and frugal heuristics that help
them survive and cope with demands. Three of these heuristics are
the recognition heuristic (what you recognise is more important than
what you do not recognise), the take-the-best heuristic (go for the option that seems the best, ignore the other options), and the takethe-
easiest heuristic (go for the option that is easier to understand).
This, however, does not explain why humans ever developed
deliberate judgement.
Dual-process model
• According to the dual-process model, probability judgements
depend on processing within two systems. One system is intuitive
and fast, whereas the other is analytical and slow. We often rely
solely on the intuitive system.
• The dual-process model has been examined with judgement problems
for which the System 1 answer is in conflict with the System 2
answer. This research indicated that people seem to be sensitive to
the conflict, even when it is not expressed in their final decisions.
New versions of the model have been proposed to account for this
finding.
Decision making involves selecting from among several possibilities.
The quality is judged on the basis of the consequences.
Omission bias
• Omission bias involves a preference for risking harm from inaction
rather than action. It is common and has been found in medical
experts and gamblers. It is due to anticipated responsibility and
regret. Some individuals use the heuristic, “Don’t just sit there. Do
something,” and show the opposite to omission bias.
Utility theory
Make choice based on benefits and cost associated w each choice
Prev assumed that ppl make decisions rationally, based on abstract, objective, economic principles
Rational decision-making = select choice that MAXIMISES EXPECTED UTILITY (value)
Gain as much value, avoid lesser value/chance
According to prospect theory, most people show loss aversion – they
are much more sensitive to potential losses than to potential gains.
The framing effect (in which decisions are influenced by irrelevant
aspects of the situation) and the sunk-cost effect (in which good
money is thrown after bad) are consistent with the theory.
Emotional factors
• Our sensitivity to potential losses occurs in part because our
emotions (especially fear) make us cautious and risk averse.
People with damage to brain regions involved in emotional processing
make more rational decisions. However, they do not make
better decisions in general than healthy people. As a matter of fact,
the reverse is true, indicating that emotions are an essential aspect
in decision making. Emotions represent the wisdom of the ages and
allow us in general to choose what is good for us and avoid what is
bad for us, as argued by the somatic marker hypothesis.
Study context
• Many decision-making studies are based on verbal answers to
hypothetical problems. These decisions may differ from the actions
people take when faced with the same dilemma in real life, as
shown in a study with mice.
Social context
• In the real world, our decision making is influenced by the social
context and by our need to justify our decisions to other people.
Decision making sometimes becomes more biased when there is
increased accountability for decisions.
Reasoning involves drawing inferences from the knowledge we
possess. Deductive reasoning involves drawing conclusions that
are definitely valid (or invalid) provided that other statements
(premises) are assumed to be true. Inductive reasoning involves
drawing general conclusions from a sequence of observations. It
does not guarantee valid conclusions, but generates hypotheses
that can be tested. Both forms of reasoning are used in scientific
research and daily life to reach conclusions.
Syllogistic reasoning
• A syllogism consists of two premises or statements followed by a
conclusion. Syllogistic reasoning is part of deductive reasoning.
Many people make errors in syllogistic reasoning because of belief
bias. This bias involves focusing on the believability of conclusions
rather than their logical validity.
Conditional reasoning
• Conditional reasoning is a second form of deductive reasoning.
Many errors are made in conditional reasoning because most
people have a limited ability to think logically and because they are influenced by irrelevant contextual information. Individuals
high in working memory capacity perform better than those of
low capacity.
Valid: Modus ponens
P1: A, B
P2: A
C: B (If A true, B is true)
Valid: Modus tollens
P1: A, B
P2: ㄱB
C: ㄱA (if B false, A is false)
Invalid: affirming the consequent,
A, B
B
Therefore A (If B is true, A is true)
Invalid: denying the antecedent
A, B
ㄱA
Therefore ㄱB (If A is false, B)
Wason card selection task
• Performance is generally very poor on the Wason card selection
task because participants rely on simple strategies, such as the
matching bias, rather than logical reasoning. Far more correct
answers are produced when deontic rules are used because such
rules increase participants’ attention to the possibility that the rule is
false. Performance is also better when participants are familiar with
the contents of the problem and are motivated to disprove the rule.
>>> modified task (Sperber & Girotto, 2002)
Use more concrete, familiar items → find task more manageable
Julio buys things from the Internet but is concerned he will be cheated. For each order, he fills out a card. On one side of the card, he indicates whether he has paid for the item ordered, while on the other side he indicates whether he has received the item.
Theories of reasoning
• According to mental model theory, we construct mental models
when reasoning. These models represent what is true and ignore
what is false (principle of truth), even when what is false is relevant
to solving the problem. Often only a single mental model is formed
even when there are other possible mental models (principleof
parsimony).
• According to dual-process theory, reasoners use simple heuristic
processes to draw inferences. Time-consuming analytic processes
are sometimes used to revise or replace the heuristic inferences.
Informal reasoning
• Informal reasoning is about the probability of inferences, not about
their truth. The content of an argument and contextual factors are
important (unlike in deductive reasoning).
• In informal reasoning, people are motivated to persuade others of
their beliefs and views. Such motivation can produce errors (e.g.,
belief bias, myside bias). However, it can also lead people to reason
better in some situations.
• According to the Bayesian model of informal reasoning, the perceived
strength of a conclusion depends on two factors: the degree
of prior belief and the strength of the evidence. Hahn and Oaksford
(2007) further noted that positive arguments are stronger than
negative ones.
• The performance of many people on reasoning problems indicates
that their reasoning is often inadequate. However, the artificiality of
many of the problems used means that human performance on
many reasoning problems may underestimate our ability to think
rationally. People spend much of their time trying to argue persuasively
rather than to seek out the truth, and this motive can lead to
errors in tasks on judgement and deductive reasoning.
At the same time, it cannot be denied that human thinking is limited
and that this has negative consequences in life. We are rarely aware
of these consequences because we fail to see the flaws in our
thinking (the Dunning–Kruger effect). Research in how these flaws
can be mediated by better information provision is ongoing.
When describing human rationality, it is better not to use a normative
framework but a pragmatic framework. Within the pragmatic framework,
Simon speaks of bounded rationality, which was extended in
the resource-rational analysis.
Intelligence and rationality
• Highly intelligent individuals perform better than those of less
intelligence on judgement and reasoning tasks. However, the effect
is often small.
• According to the tripartite model, intelligent individuals may fail to
perform well because they lack an algorithmic mind that is good
enough to override incorrect heuristic responses, or because the
algorithmic mind is not triggered by the reflective mind. Other
research has indicated that the three parts of the tripartite model