Key Terms
CogLab: Wason Selection Task; Typical Reasoning;
Risky Decisions; Decision Making: Monty Hall
12
Reasoning and
Decision Making
Some Questions We Will Consider
Do people reason logically, or do they make
errors in reasoning? (438)
How does reasoning operate in the discoveries
made by scientists? (454)
What kinds of reasoning “traps” do people
get into when reasoning and when making
decisions? (456)
How does the fact that people sometimes
feel a need to justify their decisions affect
the process by which they make these
decisions? (471)
•
•
•
•
Deductive Reasoning: Thinking Categorically
Validity and Truth in Syllogisms
How Well Can People Judge Validity?
Mental Models of Deductive Reasoning
Deductive Reasoning: Thinking Conditionally
Forms of Conditional Syllogisms
Why People Make Errors in Conditional Reasoning:
The Wason Four-Card Problem
Demonstration: Wason Four-Card Problem
Test Yourself 12.1
Inductive Reasoning: Reaching Conclusions
From Evidence
The Nature of Inductive Reasoning
The Availability Heuristic
Demonstration: Which Is More Prevalent?
The Representativeness Heuristic
Demonstration: Judging Occupations
Demonstration: Description of a Person
Demonstration: Male and Female Births
The Confirmation Bias
Culture, Cognition, and Inductive Reasoning
Demonstration: Questions About Animals
Decision Making: Choosing Among Alternatives
The Utility Approach to Decisions
Decisions Can Depend on How Choices Are Presented
Demonstration: What Would You Do?
Justification in Decision Making
The Physiology of Thinking
The Prefrontal Cortex
Neuroeconomics: The Neural Basis of Decision Making
Something to Consider: Is What Is Good
for You Also Good for Me?
Demonstration: A Personal Health Decision
Test Yourself 12.2
Chapter Summary
Think About It
If You Want to Know More
434
435Reasoning and Decision Making
What is reasoning? One definition is the process of drawing conclusions (Leighton,
2004). Another, more specific, definition is cognitive processes by which people start
with information and come to conclusions that go beyond that information (Kurtz et al., 1999).
Whatever definition we use, there is no question that reasoning is relevant to material in
most of the chapters in this book. For example, in Chapter 3, on perception, we saw that
perceiving an object can involve inference from incomplete information (Figure 12.1); in
COGNITION REASONING
Chapter 3: Perception
Chapter 6: Long-Term Memory
Chapter 10: Language
Animal is inferred from
ambiguous shapes.
Memories are constructed
from what we remember,
plus perhaps other
information.
Meaning is created by using
knowledge we obtained
earlier to help interpret
Hamlet’s statement.
“That’s an animal lurking.”
“I remember, on the first
day of class. . . .”
“To be or
not to be . . .”
“I understand what
he is saying.”
■ Figure 12.1 Some examples of processes that fit the definition of “reasoning” that we have encountered
in previous chapters.
436 Chapter 12
Chapter 6, on long-term memory, we saw that our memories of events from the past are
created by a process of construction, also from incomplete information; and in Chapter
10, on language, we saw how understanding one part of a story can depend on inferences
based on what you know has happened before.
As these examples show, reasoning is involved in a large portion of what we study in
cognitive psychology. In this chapter we are going to focus on how cognitive psychologists
have studied reasoning by focusing on two specific types: deductive reasoning and
inductive reasoning. We first consider deductive reasoning, which involves sequences
of statements called syllogisms (Figure 12.2). For example, if we know that at least a C
average is required to graduate from State U., and that Josie is graduating from State U.,
we can logically conclude that Josie had at least a C average.
We then consider inductive reasoning, in which we arrive at conclusions about
what is probably true, based on evidence. Thus, if we know that Richard attended State U.
for 4 years and that he is now the vice president of a bank, we might conclude that it is
likely that he graduated. Notice, however, that in this example, we cannot say that he
definitely graduated (maybe he never completed all the requirements, and his father, who
is president of the bank, made him a vice president). Thus, we can make definite conclusions
based on deductive reasoning and probable conclusions based on inductive reasoning.
Studying both kinds of reasoning provides insights both about how the mind works
and about everyday thinking.
• How do people reason from premises?
• How do people reason from evidence?
• How do people make choices
between alternatives?
Inductive
reasoning
Decision
making
Thinking and
the brain
Deductive
reasoning
• What areas of the brain are important
for thinking?
■ Figure 12.2
Flow diagram for
this chapter.
437Reasoning and Decision Making
We will also consider decisions, which usually involve making choices between
alternatives, and which can involve both inductive and deductive reasoning. Finally, we
will describe how the brain is involved in thinking.
Deductive Reasoning: Thinking Categorically
Aristotle is considered the father of deductive reasoning because he introduced the basic
form of deductive reasoning called the syllogism. A syllogism includes two statements,
called premises, followed by a third statement, called the conclusion. We will first consider
categorical syllogisms, in which the premises and conclusion describe the relation
between two categories by using statements that begin with all, no, or some. An example
of a categorical syllogism is:
Syllogism (1) Premise 1: All birds are animals.
Premise 2: All animals eat food.
Conclusion: Therefore, all birds eat food.
We will refer to categorical syllogisms as simply syllogisms in the discussion that follows.
Validity and Truth in Syllogisms
The word “valid” is often used in everyday conversation to mean something is true or
might be true. For example, saying “Susan has a valid point” could mean that what Susan
is saying is true, or possibly that it should be considered further. However, when used in
conjunction with categorical syllogisms, the term validity has a very specific meaning: A
syllogism is valid when its conclusion follows logically from its two premises. Conversely,
the syllogism is invalid when the conclusion does not follow from the premises.
Determining whether a particular syllogism is valid or invalid is not always easy
because there are many different combinations of premises and conclusions. Instead of
considering all of the possibilities (which would be more appropriate for a course in
logic), we will consider a few examples of valid and invalid syllogisms. The syllogism below
is the classic example of a valid syllogism. It is called Aristotle’s “perfect” syllogism
because it was introduced by Aristotle, and it is almost immediately obvious that the
conclusion follows from the two premises.
Syllogism (2) Premise 1: All A are B.
Premise 2: All B are C.
Conclusion: Therefore, all A are C.
The conclusion begins with “therefore” to indicate that it follows from the two premises.
For brevity, we will omit the therefore in the rest of the syllogisms, with the understanding
that in valid syllogisms the conclusion always follows from the premises.
Syllogisms can be stated abstractly, in terms of A, B, and C, as in syllogism 2, or
more concretely, using meaningful terms, as in syllogism 1. Notice that syllogism 1 has
438 Chapter 12
exactly the same form as syllogism 2, with A becoming “bird,” B becoming “animals,”
and C becoming “eat food.”
Although syllogisms may seem like an artificial way of studying thinking, the advantage
of using them is that we can construct syllogisms that are either valid (like
syllogisms 1 and 2, in which the conclusion follows logically from the two premises), or
not valid (the conclusion does not follow from the premises). These syllogisms can then
be used to determine how well people evaluate validity.
A basic principle of deductive reasoning states that if the two premises of a valid syllogism
are true, the syllogism’s conclusion must be true. Syllogism 1 illustrates this principle
because both premises are true, and the syllogism is logically valid. But what happens
when one or more of the premises are not true? Consider the following syllogism:
Syllogism (3) All birds are animals.
All animals have four legs.
All birds have four legs.
In this case, the second premise is not true, and the conclusion is not true. Nonetheless,
the syllogism is still valid because validity depends on the form of the syllogism,
not its content. This apparent conflict between “validity” and “truth” is often confusing,
because “valid” is sometimes used interchangeably with “true” in everyday conversation.
But remember that with regard to syllogisms, validity refers only to the logical progression
of the premises and conclusions. As we will see, it is often difficult to separate validity
and truth, and this sometimes leads to errors in reasoning. Here is an example of an invalid
syllogism that contains premises and a conclusion that could be true.
Syllogism (4) All of the students are tired.
Some tired people are irritable.
Some of the students are irritable.
If you think this syllogism is valid, consider what happens when we replace “students”
with “men” and “irritable” with “women.”
Syllogism (5) All of the men are tired.
Some tired people are women.
Some of the men are women.
Now it is clear that this syllogism is not valid because even though two premises could
be true, it is clear that the conclusion is not. Thus, sometimes it is easy to see that a syllogism
is valid, as in syllogism 1, but sometime it isn’t easy, as in syllogism 4.
How Well Can People Judge Validity?
Why are psychologists interested in whether people can judge the validity of a syllogism?
One reason is to help answer the question, “Do people think logically?” One
approach to this question was taken by early philosophers, who said that people’s minds
work logically, and so if they do make errors in judging validity it means that they were
being careless or were not paying attention.
439Reasoning and Decision Making
Another approach, which has been adopted by most cognitive psychologists, is that
logic is not necessarily built into the human mind, and so if people do make errors,
these errors tell us something about how the mind operates. Psychologists are interested
in determining what errors occur and what contributes to these errors.
People’s performance in judging syllogisms has been determined using two methods:
(1) evaluation: Present two premises and a conclusion, and ask people to indicate
whether the conclusion logically follows from the premises; (2) production: Present two
premises and ask people to indicate what conclusion logically follows from the premises,
or if no conclusion logically follows. We will focus our attention on the evaluation
task because researchers have studied deductive reasoning extensively using this
method.
When people are tested using the evaluation method, they make errors for all syllogisms
except for syllogism 2, and for some syllogisms the error rate can be as high as
70–80 percent (Gilhooly, 1988). The exact error rate depends on a number of factors,
including whether the syllogism is stated abstractly (in terms of A’s, B’s, and C’s) or in
real-world terms (for example, in terms of birds and animals).
There are many reasons that people make errors in syllogisms. We will focus on
two that have been widely studied, the atmosphere effect and the belief bias. The atmosphere
effect states that the words All, Some, and No in the premises creates an overall
“mood” or “atmosphere” that can influence the evaluation of the conclusion. According
to the atmosphere effect, two All’s generally suggests an All conclusion. One or two
No’s suggest a No conclusion; and one or two Some’s suggest a Some conclusion. If you
look back at the syllogisms we have considered so far, you can see that application of
these ideas can lead to a correct evaluation. For example, syllogism 1, which has two
All premises and an All conclusion, is valid. However, for syllogism 4, which has a Some
premise and a Some conclusion, the atmosphere effect would lead to the incorrect conclusion
that this invalid syllogism is valid. Thus, just considering the initial terms of
premises sometimes leads to correct evaluations of validity, but also leads to errors.
According to the belief bias, if a syllogism’s conclusion is true or agrees with a
person’s beliefs, this increases the likelihood that the syllogism will be judged as valid.
In addition, if the conclusion is false, this increases the likelihood that the syllogism
will be judged as invalid. The belief bias could lead to the erroneous conclusion that syllogism
4 is valid because it seems possible that “Some of the students are irritable” could
be true. Here is an example of two invalid syllogisms—both of which have the same
form, but which have conclusions that differ in believability.
Syllogism (6) No police dogs are vicious.
Some highly trained dogs are vicious.
Some police dogs are not highly trained.
Syllogism (7) No addictive drugs are inexpensive.
Some cigarettes are inexpensive.
Some addictive drugs are not cigarettes.
When people are presented with these two syllogisms, they are more likely to indicate
that syllogism 6 is invalid because they doubt the truth of the conclusion, but will
440 Chapter 12
often indicate that syllogism 7 is valid because the conclusion is believable (Evans &
Feeney, 2004).
The results of a study in which people were presented with valid and invalid syllogisms
with both believable and not-believable conclusions are shown in Figure 12.3
(Evans et al., 1983). Figure 12.3a shows the results for valid syllogisms, which, if people
judge validity correctly, should result in 100 percent “valid” responses, as indicated by
the dashed line. It is no surprise that when the conclusion is believable, 86 percent of
the judgments were “valid.” However, when the conclusion was not believable, the rates
of “valid” responses dropped to 55 percent. This decrease in “valid” responses due to
the unbelievability of the conclusion is an example of the belief bias.
Figure 12.3b shows the results for invalid syllogisms (Evans et al., 1983). In this case
the belief bias has a huge effect. Even though the syllogism is invalid (so the percent of
“valid” responses should be 0, as shown by the dashed line at the bottom of the graph),
70 percent of the responses with believable conclusions were judged to be “valid.”
The data in Figure 12.3 indicate that people are influenced by the believability of
the conclusion, and therefore they do not always follow the rules of logic.
Mental Models of Deductive Reasoning
We have presented some data on deductive reasoning, but we still haven’t considered
how people might go about determining whether a syllogism is valid or invalid. One
way is to use diagrams, such as Venn diagrams or Euler circles (see If You Want To
Know More on page 481 for more about using diagrams to determine syllogism validity).
But Phillip Johnson-Laird (1999a) wonders whether people would use these diagram
methods if they hadn’t been taught about them, and also points out that some of
these methods don’t work for some of the more complex syllogisms.
So what could people be doing? To begin a discussion of this question, JohnsonLaird
(1995) posed a problem similar to this one (try it):
Belief
bias
Believable Not
believable
50
Percentsyllogisms
judged“valid”
0
100
Believable Not
believable
(a) Valid syllogisms (b) Invalid syllogisms
Belief
bias
■ Figure 12.3 Effect
of the belief bias on
judgments of validity.
(a) Results for valid syllogisms.
(b) Results for
invalid syllogisms. The
dashed lines indicate
the percent of syllogisms
that were actually
valid. The arrows
indicate the effect of
the belief bias (Evans
et al., 1983).
441Reasoning and Decision Making
On a pool table there is a black ball directly above the cue ball. The green ball is on
the right side of the cue ball, and there is a red ball between them. If I move so the
red ball is between me and the black ball, the cue ball is to the _______ of my line of
sight.
How did you go about solving this problem? Johnson-Laird points out that the
problem can be solved by applying logical rules, but that most people solve it by imagining
the way the balls are arranged on the pool table (see Figure 12.21 on page 484). This
idea, that people will imagine situations, is the basis of Johnson-Laird’s proposal that
people use mental models to solve deductive reasoning problems.
A mental model is a specific situation that is represented in a person’s mind that
can be used to help determine the validity of syllogisms in deductive reasoning. The
basic principle behind mental models is that people create a model, or representation of
the situation, for a reasoning problem. They generate a tentative conclusion based on
this model, and then look for exceptions that might falsify the model. If they do find an
exception, they modify the model. Eventually, when they can find no more exceptions,
they accept their final model.
We can illustrate how this would work for a categorical syllogism by using the following
example (from Johnson-Laird, 1999a):
None of the artists are beekeepers.
All of the beekeepers are chemists.
Some of the chemists are not artists.
We will imagine that we are visiting a meeting of the Artists, Beekeepers, and Chemists
Society (the ABC Society, for short). We know that everyone who is eligible to be
a member must be an artist, a beekeeper, or a chemist, and that they must also abide by
the following rules, which correspond to the first two premises of the syllogism above:
No artists can be beekeepers.
All of the beekeepers must be chemists.
Our task is made easier because we can tell what professions people have by what
they are wearing. As shown in Figure 12.4, artists are wearing berets, beekeepers are
(c) Chemists(a) Artists (b) Beekeepers
■ Figure 12.4 Costumes of
the Artists, Beekeepers, and
Chemists Society (see text).
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
442 Chapter 12
wearing protective beekeeper’s veils, and chemists are wearing molecule hats. According
to the rules, no artists can be beekeepers, so people wearing berets can never wear
beekeeper’s veils. Also the fact that all beekeepers must be chemists means that everyone
wearing a beekeeper’s veil must also be wearing a molecule hat. When we meet
Alice, we know she is an artist because of her beret, and we notice she is following the
rule of not being a beekeeper (Figure 12.5a). Then we meet Beechem, who is wearing a
combination beekeeper-molecule getup, in line with the rule that all beekeepers must
be chemists (Figure 12.5b). Remember that the conclusion that has been proposed has
to do with artists and chemists. Based on what we have seen so far, we can formulate our
first model: No artists are chemists.
Remember, however, that once we have proposed our first model, we need to look
for possible exceptions that would falsify this model. We do this by milling around in
the crowd until we meet Cyart, who is wearing a beret and a molecule hat (Figure 12.5c).
We note that he is not violating the rules, so we now know that No artists are chemists
cannot be true, and, thinking back to Beechem, the beekeeper-chemist, we revise our
model to Some of the chemists are not artists.
(a) Alice
(c) Cyart
(b) Beechem
Model 1: No artists are chemists.
Model 2: Some of the chemists are not artists.
■ Figure 12.5 Costumes worn by (a) Alice (artist and not beekeeper); (b) Beechem (beekeeper and
chemist); (c) Cyart (artist and chemist).
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
443Reasoning and Decision Making
We keep looking for an exception to this rule, but find only Clara, who is a chemist,
which is also allowed by the membership regulations. But this case does not refute our
new model, and after more searching, we can’t find anyone else in the room whose existence
would refute this syllogism’s conclusion, so we accept it. This example illustrates
the basic principle behind the mental model theory: A conclusion is valid only if it cannot
be refuted by any model of the premises.
The mental model theory is attractive because it can be applied without training in
the rules of logic, and because it makes predictions that can be tested. For example, the
theory predicts that syllogisms that require more models will be more difficult to solve,
and this prediction has been confirmed in experiments (Buciarelli & Johnson-Laird,
1999).
There are other proposals about how people might test syllogisms (see Rips, 1995,
2002), but there isn’t agreement among researchers regarding the correct approach.
We have presented the mental model theory because it is supported by the results of a
number of experiments and because it is one of the models that is easiest to apply and
explain. However, a number of challenges face researchers who are trying to determine
how people evaluate syllogisms. These problems include the fact that people use a variety
of different strategies in reasoning, and that some people are much better at solving
syllogisms than others (Buciarelli & Johnson-Laird, 1999). Thus, the question of how
people go about solving syllogisms remains to be answered.
The categorical syllogisms we have been considering so far have premises and conclusions
that begin with All, Some, or No. We will now consider another type of syllogism
called conditional syllogisms, in which the first premise has the form “If . . . then. . . .”
Deductive Reasoning: Thinking Conditionally
Conditional syllogisms have two premises and a conclusion, like the ones we have
been discussing, but the first premise has the form “If . . . then. . . .” This kind of deductive
reasoning is common in everyday life. For example, let’s say that you lent your
friend Steve $20, but he has never paid you back. Knowing Steve, you might say to yourself
that you knew this would happen. Stated in the form of a syllogism, your reasoning
might look like this: If I lend Steve $20, then I won’t get it back. I lent Steve $20. Therefore,
I won’t get my $20 back.
Forms of Conditional Syllogisms
There are four major types of conditional syllogisms. Each of them is listed in Table 12.1.
They are presented in abstract form (p and q) on the left, and in concrete “everyday” form
on the right. For conditional syllogisms, the notations p and q are typically used instead
of A and B. P is the first, or “If,” term, called the antecedent, and q is the second, or
“then,” term, called the consequent.
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
444 Chapter 12
Syllogisms 1 and 2 are valid. Syllogism 1 is called affirming the antecedent because
the antecedent, p, or studying, in the second premise is affirmed. Syllogism 2 is
called denying the consequent because the consequent, q, or getting a good grade, in
the second premise is negated.
Syllogism 3 is called affirming the consequent, because q is affirmed in the second
premise. This conclusion is invalid, because even though you didn’t study, it is
still possible that you could have received a good grade. Perhaps the exam was easy,
or maybe you knew the material because it was about your job experience. If that explanation
is not convincing, consider the following syllogism, which has the same
form as syllogism 3, with “studying” replaced by “robin” and “good grade” replaced
by “bird.”
If it’s a robin, then it’s a bird.
It’s a bird.
Therefore, it’s a robin.
When stated in this way, it becomes more obvious that the affirming the consequent
form of the syllogism is invalid.
Syllogism 4 is called denying the antecedent, because p is negated (not p) in the
second premise. As in syllogism 3, you can probably think of situations that would
contradict the conclusion, in which a good grade was received even though the person
didn’t study. Again, this fact that this syllogism is invalid becomes more obvious when
restated in terms of birds and robins:
Table 12.1 Four Syllogisms That Begin With the Same First Premise
First premise of all syllogisms:
If p, then q (abstract version), or
If I study, then I’ll get a good grade (concrete example)
Is it
Syllogism Second Premise Conclusion Valid? Performance
Syllogism 1: p (Abstract) Therefore, q
Affirming the antecedent I studied (Concrete) Therefore, I’ll get Yes 97%
a good grade
Syllogism 2: Not q Therefore, not p
Denying the consequent I didn’t get a good Therefore, I didn’t Yes 60%
grade study
Syllogism 3: q Therefore, p
Affirming the consequent I got a good grade Therefore, I studied No 40%
Syllogism 4: Not p Therefore, not q
Denying the antecedent I didn’t study Therefore, I didn’t No 40%
get a good grade
Source: Performance daa from Evans et al. (1993).
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
445Reasoning and Decision Making
If it is a robin, then it’s a bird.
It’s not a robin.
Therefore, it’s not a bird.
How well can people judge the validity of these syllogisms? The results of many
experiments, shown in the far right column of Table 12.1, indicate that most people
(close to 100 percent in most experiments) correctly judge that syllogism 1 is valid, but
that performance is lower on syllogism 2, which is also valid, and 3 and 4, which are not
valid. These percentages are the average results from many studies in which the letters
p and q were used for the antecedent and the consequent. We have already seen that the
ease of determining the validity of conditional syllogisms can be greatly affected by
whether the task is stated abstractly (in terms of p’s and q’s) or concretely (studying and
grades; robins and birds). In the next section we will describe a reasoning problem that
has been studied using an abstract task and many different forms of real-world tasks.
Why People Make Errors in Conditional Reasoning:
The Wason Four-Card Problem
If reasoning from conditional syllogisms depended only on applying rules of formal
logic, then it wouldn’t matter whether the syllogism was stated in terms of abstract symbols,
such as p’s and q’s, or in terms of real-world items, such as studying or robins.
However, we know that people are often better at judging the validity of syllogisms
when real-world items are substituted for abstract symbols, and we also know that realworld
items can sometimes lead to errors, as when people are influenced by the belief
bias. Evidence for the effect of using real-world items in a conditional-reasoning problem
is provided by a series of experiments involving the Wason four-card problem.
Try this task in the following demonstration.
Demonstration
Wason Four-Card Problem
Four cards are shown in Figure 12.6. There is a letter on one side of each card and a number
on the other side. Your task is to indicate the minimum number of cards you would need to
turn over to test the following rule: If there is a vowel on one side, then there is an even number
on the other side.
Wason
Selection
Task
Wason
Selection
Task
E K 4 7
If vowel, then even number.
■ Figure 12.6 The Wason fourcard
problem (Wason, 1966).
Follow the directions in the demonstration
and try this problem.
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
446 Chapter 12
When Wason (1966) posed this task (which we will call the abstract task from now
on), 53 percent of his participants indicated that the E must be turned over. This is
correct because turning over the E directly tests the rule. (If there is an E, then there
must be an even number, so if there is an odd number on the other side, this would
prove the rule to be false.) However, another card needs to be turned over to fully test
the rule. Forty-six percent of Wason’s participants indicated that in addition to the E,
the 4 would need to be turned over. The problem with this answer is that if a vowel is
on the other side of the card, this is consistent with the rule, but if a consonant is on the
other side, turning over the 4 tells us nothing about the rule, because having a consonant
on one side and a vowel on the other does not violate the rule. As shown in Figure
12.7a, only 4 percent of Wason’s participants came up with the correct answer that
the card with the 7 also needs to be turned over. Turning over the 7 is important because
revealing a vowel would disconfirm the rule.
The key to solving the card problem is to be aware of the falsification principle:
To test a rule, it is necessary to look for situations that falsify the rule. As you can see from
Table 12.2, the only two cards that have the potential to achieve this are the E and the 7.
Thus, these are the only two cards that need to be turned over to test the rule.
The Role of “Regulations” in the Wason Task The Wason task has generated a great deal of
research. One reason for the degree of interest in this problem is because it is a conditional-reasoning
task. (Note that the problem is stated as an “If . . . then . . .” statement.)
But the main reason researchers are interested in this problem is that they want to know
why participants make so many errors.
One way researchers have gone about answering this question is to determine how
participants perform when the problem is restated in real-world terms. In one of these
50
Percentcorrect
0
100
Abstract “Beer” Abstract “Postage”
(a) (b) (c)
Abstract
■ Figure 12.7 Performance on different versions of the four-card problem. (a) Abstract version in
Figure 12.6. (b) Abstract version and the beer/drinking-age version in Figure 12.8. (c) Abstract version
and the postage version in Figure 12.9. Performance is better on concrete versions of the problem.
447Reasoning and Decision Making
experiments Richard Griggs and James Cox (1982) used the following version of the
problem:
Four cards are shown in Figure 12.8. Each card has an age on one side and the name
of a beverage on the other side. Imagine you are a police officer who is applying the
rule “If a person is drinking beer, then he or she must be over 19 years old.” (The
participants in this experiment were from Florida, where the drinking age was 19
at the time.) Which of the cards in Figure 12.8 must be turned over to determine
whether the rule is being followed?
This beer/drinking-age version of Wason’s problem is identical to the abstract version
except that concrete everyday terms (beer, soda, and ages) are substituted for the
letters and numbers. Griggs and Cox found that for this version of the problem, 73 percent
of their participants provided the correct response: It is necessary to turn over the
“beer” and the “16 years” cards. In contrast, none of their participants answered the abTable
12.2 Outcomes of Turning Over Each Card
in the Wason Task
The Rule: If there is a vowel on one side, then there is an even
number on the other side.
If turn
over . . . And the result is . . . Then this ______ the rule.
E Even confirms
E Odd falsifies
K Even is irrelevant to*
K Odd is irrelevant to
4 Vowel confirms
4 Consonant is irrelevant to
7 Vowel falsifies
7 Consonant is irrelevant to
*This outcome of turning over the card is irrelevant because the rule does
not say anything about what should be on the card if a consonant is on one
side. Similar reasoning holds for all of the other irrelevant cases.
Beer Soda
24
years
old
If drinking beer, then over 19 years old.
16
years
old
■ Figure 12.8 The beer/drinking-age
version of the four-card
problem (Griggs & Cox, 1982).
448 Chapter 12
stract task correctly (Figure 12.7b). Why is the concrete task easier than the abstract
task? Apparently, being able to relate the beer task to regulations about drinking makes
it easier to realize that the “16 years” card must be turned over.
The idea that knowing about regulations helps solve the Wason task has also been
demonstrated using the cards in Figure 12.9 and the following instructions:
Pretend you are a postal worker sorting letters. According to postal regulations, if a
letter is sealed, it must have a 5d stamp on it (d is pence in Great Britain). Which of
the four envelopes in Figure 12.9 would you have to turn over to determine whether
the rule is being obeyed?
The answer to the English postal version of the problem is that the sealed envelope
and the one with a 4d stamp on it must be turned over to check the rule. When Philip
Johnson-Laird and coworkers (1972) presented this problem to English participants
who were familiar with an actual postal regulation similar to the one stated in the problem,
performance was 81 percent correct, compared to 15 percent for the abstract task
(Figure 12.7c). When Griggs and Cox (1982) tested American participants on the postal
regulation task, they did not observe the large improvement in performance observed
for the English participants. The reason for this result appears to be that the American
participants were not familiar with postal regulations that specified different postage
for opened and sealed envelopes.
The Role of “Permissions” in the Wason Task Patricia Cheng and Keith Holyoak (1985) took
the Wason task a step further by proposing the concept of pragmatic reasoning schemas. A
pragmatic reasoning schema is a way of thinking about cause and effect in the world
that is learned as part of experiencing everyday life. One schema that people learn is the
permission schema, which states that if a person satisfies condition A (such as being
the legal age for drinking), then they get to carry out action B (being served alcohol).
The permission schema for the drinking problem “If you are 19, then you got to drink
beer” is something that most of the participants in this experiment had learned, so they
were able to apply that schema to the card task.
This idea that people apply a real-life schema like the permissions schema to the
card task makes it easier to understand the difference between the abstract version of
the card task and the “drinking beer” or “postal regulation” versions. The abstract task
5d 4d
If sealed, then 5d stamp.
■ Figure 12.9 Postage version of the four-card problem (Johnson-Laird et al., 1972).
449Reasoning and Decision Making
is set up so that participants approach it as a problem in which their goal is to indicate
whether an abstract statement about letters and numbers is true. But the drinking-beer
task and postal-regulation tasks are set up so participants approach it as problems in
which their goal is to be sure that a person has permission to drink alcohol or mail a letter.
Apparently, activating the permission schema helps people to focus attention on the
far right card, which participants often ignore for the abstract task.
To test this idea that a permissions schema may be involved in reasoning about the
card task, Cheng and Holyoak ran an experiment with two groups of participants who
both saw the cards in Figure 12.10. One of the groups was read the following directions:
You are an immigration officer at the International Airport in Manila, capital of the
Philippines. Among the documents you have to check is a sheet called Form H. One
side of this form indicates whether the passenger is entering the country or in transit,
and the other side of the form lists names of tropical diseases. You have to make
sure that if the form says “Entering” on one side, that the other side includes cholera
among the list of diseases.* Which of the following forms would you have to turn
over to check? Indicate only those that you need to check to be sure. [*The asterisk is
explained in the text that follows.]
Sixty-two percent of the participants in this group chose the correct cards (“Entering”
and “Typhoid, Hepatitis”). Participants in the other
group saw the same cards and heard the same instructions as
the first group, but with the following changes: Instead of saying
that the form listed tropical diseases, it said that the form
listed inoculations the travelers had received in the past 6 months.
In addition, the following sentence was added where indicated
by the asterisk (*): This is to ensure that entering passengers are
protected against the disease.
The changes in the instructions were calculated to achieve
a very important effect: Instead of checking just to see whether
the correct diseases are listed on the form, the immigration
officer is checking to see whether the travelers have the inoculations
necessary to give them permission to enter the country.
These instructions were intended to activate the participants’
permissions schema, and apparently this happened, because
91 percent of the participants in this condition picked the correct
cards (Figure 12.11).
■ Figure 12.10 Cholera version
of the four-card problem
(Cheng & Holyoak, 1985).Entering Transient
Cholera
Typhoid
Hepatitis
Typhoid
Hepatitis
If entering, then cholera is listed.
No permission Permission
50
Percentcorrect
0
100
■ Figure 12.11 Results of Cheng and
Holyoak’s (1985) experiment that used two
versions of the cholera problem. When
“permissions” are implied by the instructions,
performance is better.
450 Chapter 12
An Evolutionary Approach to the Four-Card Problem One of the things we have learned from
our descriptions of cognitive psychology research is that one set of data can be interpreted
in different ways by different investigators. We saw this in the case of the misinformation
effect in Chapter 7, in which memory errors were caused by presenting
misleading postevent information (MPI) after a person witnessed an event. We saw that
one group of researchers explained these errors by stating that the MPI distorted existing
memories (Loftus, 1993), but that other researchers offered explanations based on
the effect of retroactive interference and source monitoring errors (Lindsay, 1990).
Our consideration of the Wason four-card problem now leads us to another controversy,
in which different explanations have been offered to explain the results of various
experiments. For example, one proposed alternative to a permissions schema is that performance
on the Wason task is governed by a built-in cognitive program for detecting
cheating. Let’s consider the rationale behind this idea.
Leda Cosmides and John Tooby (1992) are among psychologists who have an evolutionary
perspective on cognition. They argue that we can trace many properties
of our minds to the evolutionary principles of natural selection. According to natural
selection, adaptive characteristics—characteristics that help a person or animal survive
to pass their genes to the next generation—will, over time, become basic characteristics
of humans. Charles Darwin originally proposed this theory based on observations of
physical characteristics. For example, Darwin observed that birds in a specific area had
beaks with shapes adapted to enable them to obtain the food that was available.
Applying this idea to cognition, it follows that a highly adaptive feature of the mind
would, through a similar evolutionary process, become a basic characteristic of the
mind. One such characteristic, according to the evolutionary approach, is related to
social-exchange theory, which states that an important aspect of human behavior is
the ability for two people to cooperate in a way that is beneficial to both people. Thus,
when caveman Morg lends caveman Eng his carving tool in exchange for some food
that Eng has brought back from the hunt, both people benefit from the exchange.
Everything works well in social exchange as long as each person is receiving a benefit
for whatever he or she is giving up. However, problems arise when someone cheats.
Thus, if Morg gives up his carving tool, but Eng fails to give him the food, this does
not bode well for Morg. It is essential, therefore, that people be able to detect cheating
behavior so they can avoid it. According to the evolutionary approach, people who can
do this will have a better chance of surviving, so “detecting cheating” has become a part
of the brain’s cognitive makeup.
The evolutionary approach proposes that the Wason problem can be understood in
terms of cheating. Thus, people do well in the postal version of the four-card problem
(Figure 12.9) because they can detect cheaters—someone who mails a letter with incorrect
postage. Similarly, people do well in the cholera task (Figure 12.10) because they
can detect someone who cheats by entering the country without a cholera shot.
To test the idea that cheating (and not permissions) is the important variable in
the four-card problem, Cosmides and Tooby (1992) devised a number of four-card sceFor
more Cengage Learning textbooks, visit www.cengagebrain.co.uk
451Reasoning and Decision Making
narios involving unfamiliar situations. Remember that one idea behind the permissions
schema is that people perform well because they are familiar with various rules.
To create unfamiliar situations, Cosmides and Tooby created a number of experiments
that took place in a hypothetical culture called the Kulwane. Participants in these
experiments read a story about this culture, which led to the conditional statement, “If
a man eats cassava root, then he must have a tattoo on his face.” Participants saw the
following four cards: (1) eats cassava roots; (2) eats molo nuts; (3) tattoo; and (4) no tattoo.
Their task was to determine which cards they needed to turn over to determine
whether the conditional statement above was being adhered to. This is a situation unfamiliar
to the participants, and one in which cheating could occur, because a man who
eats the cassava root without a tattoo would be cheating.
Cosmides and Tooby found that participants’ performance was high on this task,
even though it was unfamiliar. They also ran other experiments in which participants
did better for statements that involved cheating than for other statements that could not
be interpreted in this way (Cosmides, 1989; also see Gigerenzer & Hug, 1992).
However, in response to this proposal, other researchers have created scenarios
that involve unfamiliar permission rules. For example, Ken Manktelow and David Over
(1990) tested people using a rule that said, “If you clean up spilt blood, you must wear
gloves.” Note that this is a “permission” statement that most people have not heard
before, and which does not involve cheating. However, stating the problem in this way
caused an increase in performance, just like many of the other examples of the Wason
task that we have described.
The controversy continues among those who feel permissions are important, those
who focus on cheating, and researchers who have proposed other explanations for the
results of the Wason task. This is mainly because evidence has been presented for and
against every proposed mechanism (Johnson-Laird, 1999b; Manktelow, 1999).
We are left with the important finding that the context within which conditional
reasoning occurs makes a big difference. Stating the four-card problem in terms of familiar
situations can often generate better reasoning than abstract statements or statements
that people cannot relate to. However, familiarity is not always necessary for
conditional reasoning (as in the tattoo problem), and situations have also been devised
in which people’s performance is not improved, even in familiar situations (Evans &
Feeney, 2004; Griggs, 1983; Manktelow & Evans, 1979).
Sometimes controversies such as this one are frustrating to read about because, after
all, aren’t we looking for “answers”? But another way to look at controversies is that
they illustrate the complexity of the human mind and the challenge facing cognitive
psychologists. Remember that at the beginning of this book, we described an experiment
by Donders that involved simply indicating when a light was presented or whether
the light was presented on the right or on the left (see Chapter 1, page 6). We described
Donders’ experiment to illustrate the basic principle that cognitive psychologists must
infer the workings of the mind from behavioral observations. It is fitting, therefore,
that in this, the last chapter of the book, we are now describing a task that involves
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
452 Chapter 12
mental processes far more complex than judging whether a light has flashed, but which
illustrates exactly the same principle—the workings of the mind must be inferred from
behavioral observations.
We see in this controversy, about how people deal with the Wason task, how a number
of different hypotheses about what is happening in the mind can be plausibly inferred
from the same behavioral evidence. Perhaps, in the end, the actual mechanism
will be something that has yet to be proposed, or perhaps the mind, in its complexity, has
a number of different ways of approaching the Wason task, depending on the situation.
Test Yourself 12.1
1. What is deductive reasoning? What does it mean to say that the conclusion to a syllogism
is “valid”? How can a conclusion be valid but not true? True but not valid?
2. How well can people judge the validity of syllogisms? Why are psychologists interested
in the errors that people make in judging validity?
3. What is a categorical syllogism, and how do the atmosphere effect and belief bias
influence evaluation of categorical syllogisms?
4. What is a mental model for reasoning? Explain how this model can be applied to
the Artists, Beekeepers, and Chemists syllogism on page 441.
5. What is a conditional syllogism? Which of the four types of syllogisms described in
the chapter are valid, which are not valid, and how well can people judge the validity
of each type?
6. What is the Wason four-card problem, and what do the results of experiments that
have used abstract and concrete versions of the problem indicate about the roles of
(a) concreteness; (b) knowledge of regulations; and (c) permissions schemas in solving
this problem?
7. How has the evolutionary approach to cognition been applied to the Wason fourcard
problem? What can we conclude from all of the experiments on the Wason
problem?
Inductive Reasoning: Reaching Conclusions From Evidence
In deductive reasoning, premises are stated as a fact, such as “All robins are birds.”
However, in inductive reasoning, premises are based on observation of one or more specific
cases, and we generalize from these cases to a more general conclusion.
The Nature of Inductive Reasoning
For inductive reasoning, conclusions do not definitely follow from premises, and conclusions
are suggested, with varying degrees of certainty. This is illustrated by these two
inductive arguments:
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
453Reasoning and Decision Making
Observation: All the crows I’ve seen in Pittsburgh are black. When I visited my
brother in Washington, DC, the crows I saw there were black too.
Conclusion: I think it is a pretty good bet that all crows are black.
Observation: Here in Nashville, the sun has risen every morning.
Conclusion: The sun is going to rise in Nashville tomorrow.
Notice there is a certain logic to each argument, but the second argument is more
convincing than the first. Let’s consider some of the things that influence the strength
of an inductive argument.
Determining the Strength of an Inductive Argument In evaluating inductive arguments, we
do not consider validity, as we did for deductive arguments, but instead we decide how
strong the argument is. Strong arguments are more likely to result in conclusions that
are true, and weak arguments to result in conclusions that are not as likely to be true.
Remember that for inductive arguments we are dealing with what is probably true, not
what is definitely true.
There are a number of factors that can contribute to the strength of an inductive
argument. Among them are the following:
● Representativeness of observations: How well do the observations about a particular
category represent all of the members of that category? Clearly, the
crows example suffers from a lack of representativeness because it does not
consider crows from other parts of the country. If there are rare blue crows
in California, then the conclusion is not true.
● Number of observations: The argument about the crows is made stronger by
adding the Washington, DC, observations to the Pittsburgh observations.
Adding more observations would strengthen it further. The conclusion
about the sun rising in Nashville is extremely strong because it is supported
by a very large number of observations.
● Quality of the evidence: Stronger evidence results in stronger conclusions. For
example, although the conclusion that “The sun will rise in Nashville” is
extremely strong because of the number of observations, it becomes even
stronger when we consider scientific descriptions of how the earth rotates
on its axis and revolves around the sun. Thus, adding the observation that
“scientific measurements of the rotation of the earth indicate that every
time the earth rotates the sun will appear to rise” strengthens the conclusion
even further.
Bringing in scientific evidence to support an inductive argument illustrates the
connection between inductive and deductive reasoning because the scientific observation
about the rotation of the earth can be stated as the following deductive syllogism:
If the earth rotates around its axis, then the place where Nashville is located will
experience sunrise for each rotation.
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
454 Chapter 12
The earth rotates on its axis.
Therefore, Nashville will experience sunrise for each rotation.
The possibility of using scientific evidence in both inductive and deductive arguments
illustrates that although it is important to distinguish between the two types of
reasoning, the borderline between them can sometimes become fuzzy. We will describe
the link between inductive and deductive reasoning further as we consider how inductive
reasoning is used in science.
Inductive Reasoning in Science Inductive reasoning is the basic procedure used to make
scientific discoveries. The goal in science is to discover something new. To achieve this,
scientists often make systematic observations. These observations can include taking a
poll about political attitudes, observing social behavior in a shopping mall, or doing a
laboratory experiment on the Wason four-card problem. If these observations yield interesting
data, these observations can be generalized to a larger population and perhaps
be used to create a theory that goes beyond the specific observations.
In most scientific research, and especially in psychology, we base our conclusions
on more than one observation. We test a large number of participants and may run
an experiment in a number of different ways. Added participants and obtaining similar
results in variations of an experiment all strengthen our conclusions. The strength of
scientific conclusions also depends on the representativeness of our observations.
An obvious example of the importance of representative observations is determining
attitudes by polling. Predicting the presidential election by taking a poll of students
in your cognitive psychology class might tell you something about the cognitive psychology
class, but would not necessarily provide an accurate prediction of the election
results. To make a statement about attitudes in the United States it is necessary to elicit
opinions from a representative cross-section of people in the United States. Similarly,
conclusions from laboratory experiments in cognitive psychology can be safely generalized
only to the population represented in the sample of people who participate in the
experiment. Thus, it would be a mistake to say that the results for the beer/drinkingage
version of the Wason four-card experiment will necessarily generalize to people in a
society in which there are no laws regulating drinking.
Inductive reasoning is used not only to make the jump from specific observations
to more general conclusions, but to create hypotheses for further experiments. An example
of how inductive reasoning has been used to devise scientific experiments is provided
by the way Cheng and Holyoak devised the cholera experiment we described on
page 449 (Holyoak, 2003). The first step was to observe that people’s performance on
the Wason task improves when the task is stated in terms of receiving “permission” to
do something, as is the case in the beer/drinking-age and stamp/sealed-envelope versions
of the task.
Based on this observation, Cheng and Holyoak reasoned that making people more
aware that permissions are involved when they are trying to solve the Wason task should
improve their performance. They devised a way to test this idea by creating the scenario
for the cholera problem, which included one condition in which permissions were
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
455Reasoning and Decision Making
not emphasized and another condition in which they were. We can state the reasoning
behind their thinking about this experiment (which, as we have described above, was
created using inductive reasoning), by the following deductive syllogism:
If a permissions schema is activated, then performance on the Wason task should
improve.
In this experiment, one of the groups will read a sentence that will activate a permissions
schema.
Therefore, the performance of this group will improve.
As we saw when we described the “cholera” experiment, adding a sentence that was
designed to activate a permissions schema did, in fact, increase performance, thereby
confirming the “If . . . then . . .” premise of the conditional argument. Thus, inductive
reasoning can be used to generate a hypothesis about what might be going on, and deductive
reasoning sets forth the rationale of the experiment that is designed to test this
hypothesis.
Inductive Reasoning in Everyday Life Inductive reasoning is used not just for creating scientific
experiments, but for determining many of the choices we make in everyday life. For
example, Sarah has observed, from a course she took with Professor X, that he asked a
lot of questions about experimental procedures on his exams. Based on this observation,
Sarah concludes that the exam she is about to take in another of Professor X’s courses
will probably be similar. In another example, Sam has bought merchandise from mailorder
company Y before and got good service, so he places another order based on the
assumption that he will continue to get good service. Thus, anytime we make a prediction
about what will happen based on our observations about what has happened in the past,
we are using inductive reasoning.
It makes sense that we make predictions and choices based on past experience, especially
when predictions are based on familiar situations such as studying for an exam
or buying merchandise by mail. However, we make so many assumptions about the
world, based on past experience, that we are using inductive reasoning constantly, often
without even realizing it. For example, did you run a stress test on the chair you are sitting
in to be sure it wouldn’t collapse when you sat down? Probably not. You assumed,
based on your past experience with chairs, that it would not collapse. This kind of inductive
reasoning is so automatic that you are not aware that any kind of “reasoning” is
happening at all. Think about how time consuming it would be if you had to approach
every experience as if you were having it for the first time. Inductive reasoning provides
the mechanism for using past experience to guide present behavior.
When people use past experience to guide present behavior, they often use shortcuts
to help them reach conclusions rapidly. After all, we don’t have the time or energy
to stop and gather every bit of information that we need to be 100 percent certain that
every conclusion we reach is correct. These shortcuts take the form of heuristics, which
are “rules of thumb” that are likely to provide the correct answer to a problem, but
which are not foolproof.
456 Chapter 12
Using heuristics may sound familiar because we saw that people use heuristics
to help them understand what they are seeing (Chapter 3, page 79) and what a sentence
means (Chapter 10, page 373). There are a number of heuristics that people use
in reasoning that often lead to the correct conclusion, but that sometimes do not. We
will now describe two of these heuristics, the availability heuristic and the representative
heuristic.
The Availability Heuristic
The following demonstration introduces the availability heuristic.
Demonstration
Which Is More Prevalent?
Answer these questions.
● Which is more prevalent, words that begin with the letter r, or words in which
r is the third letter?
● Each item in the following list consists of two different possible causes of
death. Your task is to judge, for each pair of cases, which cause of death you
consider more likely for people in the United States. Think about this question
in this way: Imagine you randomly picked someone in the United States. Will
that person be more likely to die next year from cause A or cause B?
Cause A Cause B Most likely?
Homicide Appendicitis
Auto-train collision Drowning
Measles Smallpox
Botulism Asthma
Asthma Tornado
Appendicitis Pregnancy
When faced with a choice, we are often guided by what we remember from the
past. The availability heuristic states that events that are more easily remembered are
judged as being more probable than events that are less easily remembered (Tversky &
Kahneman, 1973). Consider, for example, the problems we posed in the demonstration.
When participants were asked to judge whether there are more words with r in the first
position or the third, 70 percent stated that there are more words that begin with r, even
though in reality there are three times more words that have r in the third position
(Tversky & Kahneman, 1973; but see also Gigerenzer & Todd, 1999).
Figure 12.12 shows the results of experiments in which participants were asked to
judge the relative prevalence of various causes of death (Lichtenstein et al., 1978). The
height of the bars indicates the percentage of participants who picked the least likely al-
457Reasoning and Decision Making
ternative. The key below the graph indicates the pairs for each bar, with the least likely
cause indicated first, followed by the more likely cause. The number in parentheses indicates
the relative frequency of the more likely cause compared to the less likely cause.
For example, the far left bar indicates that 9 percent of the participants thought it was
more likely that a person would die from appendicitis compared to homicide. However,
actual mortality data indicate that 20 times more people die from homicide than from appendicitis.
Thus, in this case, 9 percent of the participants made an incorrect judgment,
but most of the participants made an accurate judgment regarding the most likely cause
of death.
For the other causes of death, a substantial proportion of the participants misjudged
which cause is more likely. In these cases, large numbers of errors were associated with
causes that were publicized by the media. For example, 58 percent thought that more
deaths were caused by tornados than by asthma, when in reality, 20 times more people
■ Figure 12.12 Likely-causes-of-death experiment results. Pairs of “causes of death” are listed below
the graph, with the least likely cause on the left. The number in parentheses on the right indicates how
many more times more people were actually killed by the cause on the right. The bars in the graph
indicate the number of people who judged the least likely alternative in each pair as causing the most
deaths. (Adapted from Lichtenstein et al., 1978.)
1
2
3
4
5
6
Appendicitis
Auto/train collision
Smallpox
Botulism
Tornado
Pregnancy
Least Most
Homicide (20)
Drowning (5)
Measles (∞)
Asthma (920)
Asthma (20)
Appendicitis (2)
vs.
vs.
vs.
vs.
vs.
vs.
1
50
Percentpickingleast
likelycauseofdeath
0
100
25
75
2 3 4 5 6
458 Chapter 12
die from asthma than from tornados. Particularly striking is that fact that 41 percent of
the participants thought that botulism caused more deaths than asthma, even though
920 times more people die of asthma.
The explanation for these misjudgments appears linked to availability. When you
try to think of words that begin with r or that have r in the third position, it is much
easier to think of words that begin with r (run, rain, real . . .) than words that have r in
their third position (word, car, arranged . . .). When someone dies of botulism or tornados
it is front-page news, whereas deaths from asthma go virtually unnoticed by the
general public (Lichtenstein et al., 1978).
An experiment by Stuart McKelvie (1997) demonstrates the availability heuristic in
another way. McKelvie presented lists of 26 names to participants. In the “famous men”
condition, 12 of the names were famous men (Ronald Reagan, Mick Jagger) and 14 were
nonfamous women. In the “famous women” condition, 12 of the names were famous
women (Tina Turner, Beatrix Potter) and 14 were nonfamous men. When participants
were asked to estimate whether there were more males or more females in the list they
had heard, their answer was influenced by whether they had heard the male-famous list
or the female-famous list. Seventy-seven percent of the participants who had heard the
male-famous list stated that there were more males in their list (notice that there were
actually fewer), and 81 percent of the participants who had heard the female-famous
list stated that there were more females in their list. This result is consistent with the
availability heuristic, because the famous names would be more easily remembered and
would stand out when participants were asked to decide whether there were more male
or female names.
An example of operation of the availability heuristic in everyday life is the drop
in the number of people flying on commercial airlines that occurred during the year
following the 9/11 terrorist attacks. The persistent images of airplanes smashing into
the World Trade Center have led many people to avoid air travel in favor of driving
even though, according to the National Transportation Safety Board, the fatality rate is
about 500 times greater for driving than for flying in a commercial airplane. Although
factors other than thinking about safety are undoubtedly involved in the drop in air
travel, it is likely that the ready availability of images and descriptions of this disaster is
one of the factors involved.
Another example of the availability heuristic is provided by how people’s judgments
are based on correlations they observe between events. For example, you might know
from past observations that when it is cloudy and there is a certain smell in the air, it is
likely to rain later in the day. Being aware of correlations can be extremely useful. If you
have observed that your boss is more likely to grant your requests when he or she is in
a good mood, you can use this knowledge to determine the best time to ask for a raise.
Although knowledge of correlations between events can be useful, sometimes people
fall into the trap of creating illusory correlations. Illusory correlations occur when
a correlation between two events appears to exist, but in reality doesn’t exist or is much
weaker than it is assumed to be.
Illusory correlations can occur when we expect two things to be related, and so we
fool ourselves into thinking they are related even when they are not. These expectations
459Reasoning and Decision Making
often take the form of stereotypes—an oversimplified generalization about a group or
class of people that often focuses on the negative. Often, people’s stereotype about the
characteristics of a particular group leads them to pay particular attention to behaviors
associated with that stereotype, and this attention creates an illusory correlation, which
reinforces the stereotype. This is related to the availability heuristic because selective
attention to the stereotypical behaviors makes these behaviors more “available” (Chapman
& Chapman, 1969; Hamilton, 1981).
We can appreciate how illusory correlations reinforce stereotypes by considering
the stereotype that gay males are effeminate. A person who believes this stereotype
might pay particular attention to effeminate gay characters on TV programs or in movies,
and to situations in which they see a person who they know is gay acting effeminate.
Although these observations support a correlation between being gay and being effeminate,
the person has ignored the large number of cases in which gay males are not effeminate.
This may be because these cases do not stand out or because the person chooses
not to pay attention to them. Whatever the reason, selectively taking into account only
the situations that support the person’s preconceptions can create the illusion that a correlation
exists, when there may be only a weak correlation or none at all.
The Representativeness Heuristic
The representativeness heuristic is based on the idea that people often make judgments
based on how much one event resembles another event.
Making Judgments Based on Resemblances The representativeness heuristic states that
the probability that an event A comes from class B can be determined by how well A
resembles the properties of class B. To put this in more concrete terms, consider the
following demonstration.
Demonstration
Judging Occupations
We randomly pick one male from the population of the United States. That male, Robert,
wears glasses, speaks quietly, and reads a lot. Is it more likely that Robert is a librarian or a
farmer?
When Amos Tversky and Daniel Kahneman (1974) presented this question in an
experiment, more people guessed that Robert was a librarian. Apparently the description
of Robert as wearing glasses, speaking quietly, and reading a lot matched these
people’s image of a typical librarian (see “illusory conjunctions” above). Thus, they
were influenced by the representativeness heuristic into basing their judgment on how
closely they think Robert’s characteristics (which correspond to “A” in our definition of
the representativeness heuristic) match those of a “typical” librarian (“B”). However, in
doing this they were ignoring another important source of information—the base rates
Typical
Reasoning
Typical
Reasoning
460 Chapter 12
of farmers and librarians in the population. The base rate is the relative proportion of
different classes in the population. In 1972, when this experiment was carried out, there
were many more male farmers than male librarians in the United States, and this base
rate leads to the conclusion that it is much more likely that Robert is a farmer (remember
that he was randomly chosen from the population).
One reaction to the farmer–librarian problem might be that the participants might
not have been aware of the base rates for farmers and librarians, and so didn’t have the
information they needed to make a correct judgment. The effect of knowing the base
rate has been demonstrated by presenting participants with the following problem:
In a group of 100 people, there are 70 lawyers and 30 engineers. What is the chance that
if we pick one person from the group at random that the person will be an engineer?
Participants given this problem correctly guessed that there would be a 30 percent
chance of picking an engineer. However, for some participants the following description
of the person who was picked was added:
Jack is a 45-year-old man. He is married and has four children. He is generally conservative,
careful, and ambitious. He shows no interest in political and social issues
and spends most of his free time on his many hobbies, which include home carpentry,
sailing, and mathematical puzzles.
Adding this description caused participants to greatly increase their estimate of the
chances that the randomly picked person ( Jack, in this case) was an engineer. Apparently,
when only base-rate information is available, people use that information to make
their estimates. However, when any descriptive information is available, people disregard
the base-rate information, and this often causes errors in reasoning.
Making Judgments Without Considering the Conjunction Rule The following demonstration illustrates
another characteristic of the representativeness heuristic.
Demonstration
Description of a Person
Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As
a student, she was deeply concerned with issues of discrimination and social justice, and
also participated in antinuclear demonstrations. Which of the following alternatives is more
probable?
1. Linda is a bank teller.
2. Linda is a bank teller and is active in the feminist movement.
The correct answer to this problem is that it is more likely that Linda is a bank
teller, but when Tversky and Kahneman (1983) posed this problem to their participants,
85 percent picked statement 2. It is easy to see why they did this. They were influenced
by the representativeness heuristic, because the description of Linda fits people’s idea of
461Reasoning and Decision Making
a typical feminist. However, in doing this they violated the
conjunction rule, which states that the probability of a conjunction
of two events (A and B) cannot be higher than the
probability of the single constituents (A alone or B alone). For
example, the probability that Anne has a red Corvette cannot
be greater than the probability that she has a Corvette,
because the two constituents together (Corvette and red) define
a smaller number of cars than one constituent (Corvette)
alone. Similarly, there are fewer feminist bank tellers than
bank tellers, so stating that Linda is a bank teller includes the
possibility that she is a feminist bank teller (Figure 12.13).
People tend to violate the conjunction rule even when it
is clear that they understand it. The culprit is the representativeness
heuristic; in the example just cited, the participants
saw Linda’s characteristics as more representative of feminist
bank teller than bank teller.
Incorrectly Assuming That Small Samples Are Representative People also make errors in reasoning
by ignoring the importance of the size of the sample on which observations are
based. The following demonstration illustrates the effect of sample size.
Demonstration
Male and Female Births
A certain town is served by two hospitals. In the larger hospital about 45 babies are born
each day, and in the smaller hospital about 15 babies are born each day. As you know, about
50 percent of all babies are boys. However, the exact percentage varies from day to day.
Sometimes it may be higher than 50 percent, sometimes lower.
For a period of 1 year, each hospital recorded the days on which more than 60 percent of
the babies born were boys. Which hospital do you think recorded more such days?
● The larger hospital?
● The smaller hospital?
● About the same (that is, within 5 percent of each other)?
When participants were asked this question in an experiment (Tversky & Kahneman,
1974), 22 percent picked the larger hospital, 22 percent picked the smaller hospital,
and 56 percent stated that there would be no difference. The group that thought
there would be no difference was presumably assuming that the birthrate for males
and females in both hospitals would be representative of the overall birthrate for males
and females. However, the correct answer is that there would be more days with over
60 percent male births in the small hospital.
Bank tellers
Feminist
bank tellers
■ Figure 12.13 Because feminist bank tellers
are a subset of bank tellers, it is always more
likely that someone is a bank teller than a feminist
bank teller.
462 Chapter 12
We can understand why this result would occur by considering a statistical rule
called the law of large numbers, which states that the larger the number of individuals
that are randomly drawn from a population, the more representative the resulting
group will be of the entire population. Conversely, samples of small numbers of individuals
will be less representative of the population. Thus, in the hospital problem it is
more likely that the percentage of boys born on any given day will be near 50 percent
in the large hospital and farther from 50 percent in the small hospital. To make this
conclusion clear, imagine that there is a very small hospital that records only one birth
each day. Over a period of a year there will be 365 births, with about 50 percent being
boys and 50 percent being girls. However, on any given day, there will be either 100
percent boys or 100 percent girls—clearly percentages that are not representative of the
overall population. The problem for reasoning, however, is that people often assume
representativeness holds for small samples, and this results in errors in reasoning. (See
Gigerenzer & Hoffrage, 1995; Gigerenzer & Todd, 1999, for additional perspectives on
how statistical thinking and heuristics operate in reasoning.)
The Confirmation Bias
One of the major roadblocks to accurate reasoning is the confirmation bias, our tendency
to selectively look for information that conforms to our hypothesis and to overlook
information that argues against it. This effect was demonstrated by Wason (1960),
who presented participants with the following instructions:
You will be given three numbers which conform to a simple rule that I have in
mind. . . . Your aim is to discover this rule by writing down sets of three numbers
together with your reasons for your choice of them. . . . After you have written down
each set, I shall tell you whether your numbers conform to the rule or not. . . . When
you feel highly confident that you have discovered the rule, you are to write it down
and tell me what it is. (p. 131)
After Wason presented the first set of numbers, 2, 4, and 6, the participants began
creating their own sets of three numbers and receiving feedback from Wason. Note
that Wason told participants only whether their numbers fit his rule. The participants
did not find out whether their rationale was correct until they felt confident enough
to actually announce their rule. The most common initial hypothesis was “increasing
intervals of two.” Because the actual rule was “three numbers in increasing order of
magnitude,” the rule “increasing intervals of two” is incorrect even though it creates
sequences that satisfy Wason’s rule.
The secret to determining the correct rule is to try to create sequences that don’t
satisfy the person’s current hypothesis, but which do satisfy Wason’s rule. Thus, determining
that the sequence 2, 4, 5 is correct, allows us to reject our “increasing intervals
of two” hypothesis and formulate a new one. The few participants whose rule was correct
on their first guess followed the strategy of testing a number of hypotheses by creating
sequences that were designed to disconfirm their current hypothesis. In contrast,
463Reasoning and Decision Making
participants who didn’t guess the rule correctly on their first try tended to keep creating
sequences that confirmed their current hypothesis.
The confirmation bias acts like a pair of blinders—we see the world according to
rules we think are correct and are never dissuaded from this view because we seek out
only evidence that confirms our rule. The confirmation bias is so strong that it can affect
people’s reasoning by causing them to ignore relevant information. Charles Lord
and coworkers (1979) demonstrated this in an experiment that tested how people’s attitudes
are affected by exposure to evidence that contradicts the attitudes.
By means of a questionnaire, Lord identified one group of participants in favor of
capital punishment and another group against it. Each participant was then presented
with descriptions of research studies on capital punishment. Some of the studies provided
evidence that capital punishment had a deterrent effect on murder; others provided
evidence that capital punishment had no deterrent effect. When the participants
reacted to the studies, their responses reflected the attitudes they had at the beginning
of the experiment. For example, an article presenting evidence that supported the deterrence
effect of capital punishment was rated as “convincing” by proponents of capital
punishment and “unconvincing” by those against capital punishment. This is the confirmation
bias at work—people’s prior beliefs caused them to focus only on information
that agreed with their beliefs and to disregard information that didn’t.
Culture, Cognition, and Inductive Reasoning
When we discussed categories in Chapter 8, we described research comparing the
way Itza and American participants think about categories (see Chapter 8, page 315, for
a description of the Itza). We saw that Itza and American participants have different
ideas about which level of categories is “basic,” with Itza focusing on the level “sparrow,
oak” and Americans focusing on the level “bird, tree.” We also saw that there are
similarities, with both American and Itza using categories similarly for an inductivereasoning
task.
We now return to these two groups to consider some further comparisons of how
they use categories in reasoning. Let’s start with a demonstration.
Demonstration
Questions About Animals
The following two questions are about animals that live on an island.
Question 1:
Porcupines have a disease.
Squirrels have another disease.
Do you think all other mammals on the island have the disease of porcupines
or the disease of squirrels?
464 Chapter 12
Question 2:
Wolves and deer have a disease.
Wolves and coyotes have another disease.
Do you think all other mammals on the island have the disease of wolves and deer
or of wolves and coyotes?
The preceding questions both involve induction, because they require reasoning
from specific observations to more general conclusions, and because the answer is a
“probably” answer rather than a “definitely” answer. These items are based on a model
of reasoning about categories called the similarity-coverage model (Osherson et al.,
1990). The goal of this model is to explain how people’s conceptions of different categories
influence the strength of inductive arguments. We will discuss this model by
posing a few more problems, and will then return to the questions from the demonstration.
One principle of the model, called the typicality principle, is illustrated by asking
which of the following is most likely to be true.
Premise: Robins have a higher potassium concentration in their blood than
humans.
Conclusion 1: Therefore, all birds have a higher potassium concentration in their
blood than humans.
Premise: Penguins have a higher potassium concentration in their blood
than humans.
Conclusion 2: Therefore, all birds have a higher potassium concentration in their
blood than humans.
According to the typicality principle, the argument with the most typical example of a
category in the premise is the strongest argument. Thus, if people think robins are more
typical of birds than penguins, they will pick conclusion 1 as more likely to be true.
Here is an example that illustrates another principle. Which of the following arguments
are more likely to be true?
Premise: Hippopotamuses have an ulnar artery.
Hamsters have an ulnar artery.
Conclusion 1: All mammals have an ulnar artery.
Premise: Hippopotamuses have an ulnar artery.
Rhinoceroses have an ulnar artery.
Conclusion 2: All mammals have an ulnar artery.
According to the diversity principle, the argument with the greatest coverage of a category
is stronger. Because hippopotamus and hamster taken together are more diverse
than hippopotamus and rhinoceros, they have higher coverage of the category mammal.
Therefore, according to the principle of diversity, conclusion 1 is stronger.
Now return to the demonstration “Questions About Animals,” and decide which argument
illustrates typicality and which illustrates diversity. (Note that the questions in
the demonstration are stated differently from the examples we have just considered, but
465Reasoning and Decision Making
it is still possible to determine which principles apply. Stop and decide, before reading
further.)
When Alejandro Lopez and coworkers (1997) presented problems like question 1 in
the demonstration to Itza and U.S. participants, they found that the overall responses
of both groups agreed with the principle of typicality (Figure 12.14a; picking squirrel
rather than porcupine). Thus, both groups’ answers reflected the animals in the premises
that they thought were more typical.
Although the results for typicality were similar for both groups, the results for diversity
were different. Thus, when the two groups were presented with problems like
question 2 in the demonstration, almost all of the U.S. participants (96 percent) gave an
answer that corresponded to the principle of diversity (picking wolves and deer), whereas
only 38 percent of the Itza participants gave an answer that corresponded to this principle
(Figure 12.14b).
What does this result mean? Why do Itza ignore diversity in making their choices?
One reason is that the Itza, but not the U.S. participants, are influenced by ecological
considerations. For example, one Itza participant was presented with the following
problem.
Rats and pocket mice have a disease. Tapirs and squirrels have a disease. Do you
think all other mammals on the island would have the disease of rats and pocket mice
or the disease of tapirs and squirrels?
The person chose “rats and pocket mice,” even though the principle of diversity
predicts that tapirs and squirrels, which are more different, would be chosen. The person
explained that tapirs and squirrels are less likely to pass on the disease because they
probably got it from another agent, such as a bat biting them, whereas rats and pocket
mice are similar enough so they don’t need an agent like a bat biting them to both get
the disease. Thus, the Itza participant was not being illogical. Rather, she was using
logic that was based on her knowledge of the animals in their environment.
American Itza
50
Percentmatchingprediction
0
100
American Itza
(a) Typicality (b) Diversity
■ Figure 12.14
(a) Percentage of
judgments of inductive
arguments that follow
the principle of typicality
for American and for
Itza participants. (b) Percentage
of judgments
that follow the principle
of diversity for American
and for Itza participants
(Lopez et al., 1997).
466 Chapter 12
Interestingly, similar effects of knowledge on reasoning about categories have been
observed for different groups of U.S. participants. For example, when asked to sort trees
into different categories, taxonomists (who are schooled in scientific biology) sorted the
trees by their scientific biological category, whereas park maintenance workers (who
are responsible for maintaining the trees) sorted the trees based on their experience in
taking care of trees. When tested, the park maintenance workers rejected the diversity
principle in favor of ecologically based reasoning—just like the Itza (Medin et al.,
1997)!
These cross-cultural and cross-occupational studies emphasize that in studying
cognition, we can’t assume that people always think in exactly the same way or use the
same information to support their thinking.
Decision Making: Choosing Among Alternatives
We make decisions every day, from relatively unimportant ones (what clothes to wear,
what movie to see) to those that can have great impact on our lives (what college to
attend, whom to marry, what job to choose). The process of decision making can involve
both inductive and deductive reasoning, so we have already considered some of
the principles that apply to the study of how people make decisions.
When we discussed the availability and representativeness heuristics we used examples
in which people were asked to make judgments about things like causes of death
or people’s occupations. As we discuss decision making, our emphasis will be on how
people make judgments that involve choices between different courses of action. These
choices can be concerned with personal decisions, such as deciding what school to attend
or whether to fly or drive to a destination, or they can be concerned with decisions
that a person might make in conjunction with their profession, such as “Which
advertising campaign should my company run?” or “What is the best economic policy
for the United States?” We begin by considering one of the basic properties of decision
making: Decisions involve both benefits and costs.
The Utility Approach to Decisions
Much of the early theorizing on decision making was influenced by economic utility
theory, which is based on the assumption that people are basically rational, so if they
have all of the relevant information, they will make a decision which results in the maximum
expected utility. Utility refers to outcomes that are desirable because they are in
the person’s best interest (Manktelow, 1999; Reber, 1995). The economists who studied
decision making thought about utility in terms of monetary value, with the goal of good
decision making being to make choices that resulted in the maximum monetary payoff.
One of the advantages of the utility approach is that it specifies procedures that
make it possible to determine which choice would result in the highest monetary value.
For example, if we know the odds of winning when playing a slot machine in a casino,
467Reasoning and Decision Making
and also know the cost of playing and the size of the payoff, it is possible to determine
that, in the long run, playing slot machines is a losing proposition. Thus, in terms of
monetary payoff, it would be unwise to play the slots.
A Problem for the Utility Approach: People Do Not Necessarily Act to Maximize Monetary Value Even
though most people realize that in the long run the casino wins, the huge popularity of
casinos indicates that many people have decided to patronize casinos anyway. Observations
such as this, as well as the results of many experiments, have led psychologists to
conclude that people do not always make decisions that maximize their monetary outcome.
This does not necessarily mean that people are irrational, but that they find value
in things other than money. Thus, for some people the fun of gambling might outweigh
the probable loss of some money, and, of course, there’s the thrill of thinking about the
possibility of “beating the odds” and being the one who hits the jackpot.
Another problem with the utility approach is that many decisions do not involve
payoffs that can be calculated. As the popular television advertisement for a credit card
says, Tickets to the ball game, $60; Hot dogs, $10; Your team’s baseball cap, $20; Seeing the
game with your son or daughter, “priceless.” Thus, utility is not always reducible to dollars
and cents, but is often in the mind of the person.
The idea that utility can be in the person’s mind brings up another potential problem
with the utility approach. When people have to make decisions that will affect their lives,
they often create mental simulations, which can sometimes be misleading (Kahneman &
Tversky, 1982; Dunning & Parpal, 1989). Mental simulations are models that people
create about what will happen following different decisions. For example, if Roberta is
trying to decide whether to go to “University A” or “University B,” she may imagine
what it would be like attending each school. In doing this she may imagine life at University
A to be intellectual and the atmosphere at University B to be more socially oriented.
Although the procedure of creating mental simulations can be useful, there is a danger
that it may not lead to accurate predictions. After all, Roberta hasn’t actually attended
either school and has no experience with college, so she is just guessing what
each school would be like. In fact, people often make inaccurate predictions about what
will happen in a particular situation. For example, when people win the lottery, they
may initially see nothing but positive outcomes, such as being able to quit their job, buy
a new house, and finance their children’s college educations. Later, however, they become
aware of negative aspects, such as being hounded by other people who want a piece
of the action, lack of privacy, losing friends, and worries about what’s happening to their
investments in the stock market. Events that people imagine will occur are often different
from the events that actually do occur (T. D. Wilson et al., 2000).
People Are Often Not Good at Predicting Their Emotional Reactions T. D. Wilson and coworkers
(2000) point out that even if people were able to accurately predict what would happen,
they are often poor at predicting how happy or unhappy the event will cause them to feel.
One of the things responsible for this lack of accuracy in predicting their emotions is
called the focusing illusion, which occurs when people focus their attention on just one
468 Chapter 12
aspect of a situation and ignore other aspects of a situation that may be important. For example,
when college students were asked the questions “How happy are you?” and “How
many dates did you have last month?” their answers depended on the order in which the
questions were asked. When the happiness question was asked first, the correlation between
the answers to the two questions was 0.12, but when the dating question was asked
first, the correlation rose to 0.66. Apparently, asking the dating question first caused participants
to focus on dating as being an important determinant of happiness, and so they
rated themselves as happier if they had a large number of dates (Strack et al., 1988).
The focusing illusion has also been demonstrated in a study that considered people’s
perceptions of how satisfied a target person would be if they lived in different locations.
The participants for this study were students at two Midwestern universities (the
University of Michigan and the University of Ohio) and at two California universities
(the University of California at Irvine and UCLA).
There were two groups, the self group and the other group. The self group rated
themselves on overall life satisfaction. The other group predicted how a hypothetical target
person who was similar to themselves would rate their life satisfaction if they lived
in California and if they lived in the Midwest. (Both groups also rated other things as
well, but for our purposes we will focus on overall life satisfaction.)
The results of this study showed that there was no difference in how the California
and Midwest students in the self group rated their own overall life satisfaction, but both
California and Midwest students in the other group predicted that their hypothetical target
person would be happier in California (Table 12.3; Schkade & Kahneman, 1998).
Why did both California and Midwest students in the other group predict that
their target person would be happier in California? The experimenters suggest that
the higher ratings for California were probably caused by the participants’ tendency
to focus on the most easily observed and distinctive differences between the two locations,
such as good weather and natural beauty (which people generally associate with
California), and to ignore other factors, such as job prospects, academic opportunities,
and financial situation. The message here is that before you decide that moving to California
will make you happier, it is important to consider a wide range of outcomes from
this decision—not just that the weather will be better.
All of the evidence above shows that people often make decisions that do not result
in maximizing monetary value. In addition, they may not be able to accurately predict
Table 12.3 Focusing Illusion Experiment
Group Question Asked Answer
Self How would you rate your life satisfaction? No difference in self-ratings for California
and Midwest students.
Other Would a person be happier living in Both California and Midwest students
California or the Midwest? answered that the hypothetical person
would have higher life satisfaction in
California.
469Reasoning and Decision Making
what outcome a particular decision will bring or how they will feel about the outcome
when it happens. We will now consider research on decision making that has shown
that people’s evaluation of different choices can depend on the way these choices are
presented.
Decisions Can Depend on How Choices Are Presented
Our discussion of deductive and inductive reasoning has shown that reasoning is affected
by more than just the facts of the situation. This also happens in decision making,
when a person’s judgments are affected by the way choices are stated. For example,
take the decision about whether to become a potential organ donor. Although a poll has
found that 85 percent of Americans approve of organ donation, only 28 percent have
granted permission by signing a donor card. This signing of the card is called the optin
procedure, because it involves the person taking an active step (Johnson & Goldstein,
2003).
The low American consent rate for organ donation also occurs in other countries,
such as Denmark (4 percent), the United Kingdom (27 percent), and Germany (12 percent).
One thing that these countries have in common is that they all use the opt-in
procedure. However, in France and Belgium the consent rate is over 99 percent. These
countries use the opt-out procedure, in which the person is a potential organ donor
unless he or she requests not to be.
Besides having important ramifications for public health (in 1995 more than 45,000
people in the United States died waiting for a suitable donor organ), the difference between
opt-in and opt-out procedures has important implications for the theory of decision
making, because according to the utility approach, people make their decisions
based on expected utility value. According to this approach, people’s decisions shouldn’t
depend on how the potential choices are stated. However, the opt-in versus opt-out results
indicates that the procedure used to identify people’s willingness to be organ donors
does have an effect.
But what about when people are confronted with hypothetical situations in which
they are forced to choose between two alternatives? The following demonstration provides
an example of such a situation.
Demonstration
What Would You Do?
Imagine that the United States is preparing for the outbreak of an unusual Asian disease
that is expected to kill 600 people. Two alternative programs to combat the disease have
been proposed. Assume that the exact scientific estimates of the consequences of the programs
are as follows:
● If Program A is adopted, 200 people will be saved.
● If Program B is adopted, there is a 1/3 probability that 600 people will be
saved, and a 2/3 probability that no people will be saved.
Risky
Decisions
Risky
Decisions
Decision
Making
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
470 Chapter 12
200 saved
400 die
1/3 chance 2/3 chance
600
saved
None
saved
Program A (72%)
Program C (22%)
Program B (28%)
Program D (78%)
1/3 chance 2/3 chance
None
die
600
die
■ Figure 12.15 How framing affects decision making. These pie charts diagram the conditions set
forth for Programs A, B, C, and D in the text. Note that the number of deaths and probabilities for programs
A and B are exactly the same as for programs C and D. The percentages indicate the percentage
of participants who picked each program when given choices between A and B or between C and
D (Tversky & Kahneman, 1981).
Which of the two programs would you favor?
Now consider the following additional proposals for combating the same disease:
● If Program C is adopted, 400 people will die.
● If Program D is adopted, there is a 1/3 probability that nobody will die,
and a 2/3 probability that 600 people will die.
Which of these two programs would you pick?
For the first pair of proposals, Program A was chosen by 72 percent of the students
in an experiment by Tversky and Kahneman (1981) and the rest picked Program B (Figure
12.15). The choice of Program A represents a risk-aversion strategy. The idea of
saving 200 lives with certainty is more attractive than the risk that no one will be saved.
However, when Tversky and Kahneman presented the descriptions of Programs C and D
to another group of students, 22 percent picked Program C and 78 percent picked Program
D. This represents a risk-taking strategy. The certain death of 400 people is less
acceptable than a 2 in 3 chance that 600 people will die.
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
471Reasoning and Decision Making
Tversky and Kahneman concluded that, in general, when a choice is framed in
terms of gains (as in the first problem, which is stated in terms of saving lives) people
use a risk-aversion strategy, and when a choice is framed in terms of losses (as in the second
problem, which is stated in terms of losing lives), people use a risk-taking strategy.
But if we look at the four programs closely, we can see that they are identical pairs
(Figure 12.15). Programs A and C both result in 200 people living and 400 people dying.
Yet 72 percent of the participants picked program A and only 22 percent picked
program C. A similar situation occurs if we compare programs B and D. Both lead to
the same number of deaths, yet one was picked by 28 percent of the participants and the
other by 78 percent. These results illustrate the framing effect—decisions are influenced
by how a decision is stated, or framed.
One reason people’s decisions are affected by framing is that the way a problem is
stated can highlight some feature of the situation (for example, that people will die) while
deemphasizing other features (Kahneman, 2003). It should not be a surprise that the way
a choice is stated can influence cognitive processes, because this is similar to what happens
when the way a syllogism is stated influences a person’s ability to determine whether
the syllogism is valid (page 438). We also saw, in the chapter on problem solving, that the
way a problem is stated can influence our ability to solve the problem (page 408).
Justification in Decision Making
To end our consideration of decision making, we will consider yet another factor that
influences how people make decisions. This factor is the need to justify the decision.
We can illustrate this by considering an experiment by Tversky and Eldar Shafir (1992),
in which they presented the following problem to two groups of students. The “pass”
group read the words that are underlined, and the “fail” group had these words replaced
by the words in italics (which the pass group did not see).
Imagine that you have just taken a tough qualifying examination. It is the end of the
semester, you feel tired and run-down, and you find out that you passed the exam/
[ failed the exam. You will have to take it again in a couple of months—after the Christmas
holidays]. You now have the opportunity to buy a very attractive 5-day Christmas
vacation package to Hawaii at an exceptionally low price. The special offer expires
tomorrow. Would you
● buy the vacation package?
● not buy the vacation package?
● pay a $5 nonrefundable fee in order to retain the rights to buy the vacation package
at the same exceptional price the day after tomorrow?
The results for the two groups are shown in the columns headed “passed” and
“failed” in Table 12.4. Notice that there is no difference between the two groups. Fiftyfour
percent of the participants in the “pass” group opt to buy the vacation package, and
57 percent of the participants in the “fail” group opt for the package. The interesting
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
472 Chapter 12
result happened when a third group received the same problem as above, except these
participants were told that the outcome of the exam wouldn’t be available for 2 more
days. Notice that only 32 percent of these participants opted for the package and that
61 percent decided they would pay the $5 so they could put off making the decision until
they knew whether they had passed or failed the exam.
Apparently what happened in this experiment is that 61 percent of the participants
in the “no result” group did not want to make a decision about the trip until they found
out whether they passed or failed, even though the results for the other two groups
indicates that passing or failing actually made no difference in the decision about the
vacation packages.
To explain this result, Tversky and Shafir suggest that once the students know the
outcome they can then assign a reason for deciding to buy the vacation. Participants who
passed could see the vacation as a reward; participants who failed could see the vacation
as a consolation that would provide time to recuperate before taking the exam again.
Although there are other possible interpretations for these results, there is a great
deal of other evidence that the decision-making process often includes looking for justification
so the person can state a rationale for his or her decision. This is why doctors
may carry out medical tests that might not lead to different treatments but that provide
additional evidence for the treatment they have recommended, thereby making it easier
to justify the treatment to themselves, their patients, and, if necessary, to the courts
(Tversky & Shafir, 1992).
The Physiology of Thinking
In this section we will consider the types of thinking we have discussed in this chapter
and the previous one. We begin by asking the question, “How is the brain involved in
problem solving, reasoning, and making decisions?” Because all of these forms of thinking
involve a number of different cognitive capacities—including perception, memory,
and the ability to focus and maintain attention—it isn’t surprising that a number of different
areas of the brain are involved. However, we will focus on one area in particular,
the prefrontal cortex, because it plays such a major role in thinking.
Table 12.4 Choice Behavior and Knowledge of Exam Outcome
Test Results
Failed Not Available
Passed Test Test for 2 Days
Buy vacation package 54% 57% 32%
Don’t buy 16 12 7
$5 to keep open option to buy later 30 31 61
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
473Reasoning and Decision Making
The Prefrontal Cortex
The prefrontal cortex (PFC; Figure 12.16) is activated by stimuli from all of the senses,
by the retrieval of memories and the anticipation of future events, and can be affected by
a person’s emotional state (Wallis et al., 2001). It is not surprising, therefore, that the PFC
plays a central role in determining complex behaviors that are involved in thinking.
One of the earliest reports of the effect of frontal lobe damage on functioning was
the case of a young homemaker who had a tumor in her frontal lobe that made it impossible
for her to plan a family meal, even though she was capable of cooking the individual
dishes (Penfield & Evans, 1935). Results such as this led to the conclusion that
the PFC plays an important role planning future activities (Owen et al., 1990).
The prefrontal cortex has been linked to problem solving in a number of ways.
Damage to the PFC interferes with people’s ability to act with flexibility, a key requirement
for solving problems. One symptom of PFC damage is a behavior called perseveration,
in which patients have difficulty in switching from one pattern of behavior
to another (Hauser, 1999; Munakata et al., 2003). For example, patients with damage to
the PFC have difficulty when the rules change in a card-sorting task. Thus, if they begin
by successfully separating out the blue cards from a pack, they continue picking the
blue cards even after the experimenter tells them to shift to separating out the brown
cards. Clearly, perseveration would play havoc with attempts to solve complex problems
for which it is necessary to consider one possible solution and then shift to another possibility
if the first one doesn’t work.
Because damage to the PFC results in perseveration and poor planning ability, it
is not surprising that PFC damage decreases performance on tasks such as the Tower
of London problem (a task similar to the Tower of Hanoi problem that involves moving
colored beads between two vertical rods; Carlin et al., 2000; Owen et al., 1990),
Frontal lobe
Prefrontal cortex
■ Figure 12.16
Brain showing location
of the prefrontal
cortex (PFC).
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
474 Chapter 12
Easy task
(a)
Hard task
(b)
100
75
50
25
Percentcorrect
0
Patients with prefrontal
cortex damage
Patients with temporal
lobe damage
Participants without
brain damage
■ Figure 12.17 Effect of damage to
the PFC on performance on a reasoning
task. Participants without brain
damage, participants with temporal
lobe damage, and participants with
PFC damage can all solve the easy
task (left bars), but the PFC group’s
performance drops to a low level
when the task is made more difficult
(Waltz et al., 1999).
the Tower of Hanoi problem (Morris et al., 1997), and the Luchins water-jug problem
(Colvin et al., 2001). Brain imaging has also shown that problem solving activates the
PFC in normal participants (Rowe et al., 2001).
Other research has shown that the PFC is important for a number of cognitive
tasks involving planning, reasoning, and making connections among different parts of
a problem or a story. For example, when Tiziana Zalla and coworkers (2002) tested patients
with PFC damage, she found that these patients were able to understand individual
words and could identify events described in stories. However, they were unable
to follow the order of events in the story or to make inferences that connected different
parts of the story.
There is also a large amount of evidence that the PFC is important for reasoning.
This has been demonstrated by presenting a deductive-reasoning task to people with
PFC damage. Participants were presented with relationships such as the following: Sam
is taller than Nate; Nate is taller than Roger, and their task was to arrange the names in
order of the people’s heights. When James Waltz and coworkers (1999) presented these
tasks to patients with PFC damage and control groups of participants without brain
damage and patients with temporal lobe damage, they found that all of these groups
did well when the task was easy, like the previous one about Sam, Nate, and Roger (Figure
12.17a). However, when the task was made more difficult by scrambling the order of
presentation (example: Beth is taller than Tina; Amy is taller than Beth), then the people
without brain damage and the patients with temporal lobe damage still did well, but the
PFC patients performed poorly (Figure 12.17b). This result confirms the conclusion of
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
475Reasoning and Decision Making
brain imaging studies, which show that as reasoning problems become more complex,
reasoning activates larger areas of the PFC (Kroger et al., 2002). (Also see “If You Want
to Know More: Neurons That Respond to Abstract Rules,” on page 483.)
Neuroeconomics: The Neural Basis of Decision Making
A new approach to studying decision making, called neuroeconomics, combines approaches
from the fields of psychology, neuroscience, and economics (Lee, 2006;
Sanfey et al., 2006). One outcome of this approach has been research that has identified
areas of the brain that are activated as people make decisions while playing economic
games. This research shows that decisions are often influenced by emotions, and that
these emotions are associated with activity in specific areas of the brain.
To illustrate the neuroeconomic approach, we will describe an experiment by Alan
Sanfey and coworkers (2003) in which people’s brain activity was measured as they
played the ultimatum game. The ultimatum game is very simple. Two people play:
One is designated as the proposer and one as the responder. The proposer is given a sum
of money, say $10, and makes an offer to the responder as to how this money should be
split between them. If the responder accepts the offer, then the money is split according
to the proposal. If the responder rejects the offer, neither player receives anything.
Either way, the game is over after the responder makes his or her decision.
According to utility theory, the responder should accept the proposer’s offer, no
matter what it is. This is the rational response, because if you accept the offer you get
something, but if you refuse, you get nothing (remember that the game is only one-trial
long, so there is no second chance).
In Sanfey’s experiment, participants played 20 separate games as responder: 10 with
10 different human partners and 10 with a computer partner. The offers made by both
the human and computer partners were determined by the experimenters, with some
being “fair” (evenly split, so the responder received $5) and some “unfair” (the responder
received $1, $2, or $3). The results of responders’ interactions with their human partners
(gray bars in Figure 12.18) match the results of other research on the ultimatum
game—all responders accept an offer of $5, most accept the $3 offer, and half or more
reject the $1 or $2 offers.
Why do people reject low offers? When Sanfey and coworkers asked participants,
many explained that they were angry because they felt the offers were unfair. Consistent
with this explanation, when participants received exactly the same offers from their computer
partner, more accepted “unfair” proposals (blue bars in Figure 12.18). Apparently,
people are less likely to get angry with an unfair computer than with an unfair person.
In addition to testing people’s behavior, Sanfey and coworkers measured brain activity
in the responders as they were making their decisions. The results showed that
the right anterior insula, an area located deep within the brain between the parietal and
temporal lobes, was activated about three times more strongly when responders rejected
an offer than when they accepted it (Figure 12.19a). Also, participants with higher activation
to unfair offers rejected a higher proportion of the offers. The fact that the
insula responded during rejection is not surprising when we consider that this area of
For more Cengage Learning textbooks, visit www.cengagebrain.co.uk
476 Chapter 12
the brain is connected with negative emotional states, including pain,
distress, hunger, anger, and disgust.
What about the prefrontal cortex, which plays such a large role in
complex cognitive behaviors? The PFC is also activated by the decision
task, but this activation is the same for offers that are rejected and
offers that are accepted (Figure 12.19b). Sanfey hypothesizes that the
function of the PFC may be to deal with the cognitive demands of the
task, which involves the goal of accumulating as much money as possible.
Looked at in this way, each of these brain areas represents a different
goal of the ultimatum game—the emotional goal of resenting
unfairness is handled by the anterior insula, and the cognitive goal of
accumulating money is handled by the PFC.
The results of this experiment support the idea that it is important
to take emotional factors into account when considering decision making.
It also illustrates the value of combining both physiological and
behavioral approaches to the study of decision making.
Something to Consider
Is What Is Good for You Also Good for Me?
When we discussed how framing affects decision making, we saw that people’s decisions
regarding programs to deal with the outbreak of a hypothetical Asian disease depended
on how the problem was stated (pages 469–470). We now pose a similar type of medical
problem, but in a more personal way, because the hypothetical decision you are asked to
make could affect you personally.
Reject
Accept
Contrastvalues
1.0
0.8
0.6
0.4
0.2
0
(a) Anterior
insula
(b) PFC
■ Figure 12.19 Response of the
insula and PFC to “fair” and “unfair”
offers (Sanfey et al., 2006).
Computer
Human
Acceptancerates
Offer
100
90
80
70
60
50
40
30
20
10
0
$7:$3 $8:$2 $9:$1$5:$5
■ Figure 12.18 Behavioral results of
Sanfey and coworkers’ (2003) experiment,
showing responders’ acceptance
rates in response to different offers
made by human partners and computer
partners.
477Reasoning and Decision Making
Demonstration
A Personal Health Decision
Imagine that there will be a deadly flu going around your area next winter. Your doctor says
that you have a 10 percent chance (10 out of 100) of dying from this flu. A new flu vaccine
has been developed and tested. If administered, the vaccine will prevent you from catching
the deadly flu. However, there is one serious risk involved: The vaccine is made from a somewhat
weaker type of flu virus, so there is a 5 percent risk (5 out of 100) that the vaccine
could kill you. Considering this information, decide between the following two alternatives:
● I will not take the vaccine, and I accept the 10 percent chance of dying from
this flu.
● I will take the vaccine, and I accept the 5 percent chance of dying from the
weaker flu in the vaccine. (Adapted from Zikmund-Fisher et al., 2006)
When Brian Zikmund-Fisher and coworkers (2006) gave this choice to their participants,
48 percent said they would take the vaccine. This is an interesting result,
because it means that 52 percent of the participants decided to do nothing, even though
statistically this doubled their chances of dying.
This result is an example of the omission bias—the tendency to do nothing to
avoid having to make a decision that could be interpreted as causing harm. However,
Zikmund-Fisher’s experiment asked participants not only to imagine that they were
making a decision for themselves, as in the demonstration, but to make the decision
while imagining themselves in the following three roles: (1) as a physician recommending
a treatment for a patient; (2) as a hospital medical director setting treatment guidelines
for all patients in the hospital; and (3) as a parent of a child who might receive the
treatment. The results of this experiment, shown in Figure 12.20, indicate that people
100
50
Percentofparticpants
choosingflushot
0
For patient
(role: physician)
For hospital
(role: medical
director)
For child
(role: parent)
For self
■ Figure 12.20 Effect
of imagined decisionmaking
role on the willingness
to choose the
flu vaccine. (Data from
Zikmund-Fisher et al.,
2006).
478 Chapter 12
are more likely to recommend that others receive the shot than they are to choose the
shot for themselves.
Apparently, the decision a person makes can be influenced by the person or group
for whom they are making the decision. But why does this occur? Zikmund-Fisher and
coworkers propose that when making decisions for others, people take into account the
possibility that they will be held responsible if something bad happens. Looked at from
this point of view, it is easy to understand why a medical director would be prone to recommend
that hospital patients receive the vaccine, because it is easy to justify a decision
that maximizes survival chances for a group of people.
The most important implication of these results may be what it suggests about
how physicians should present choices to their patients. Physicians often feel that they
should simply present the information and let their patients deal with making the decision.
But perhaps physicians should be sensitive to some of the emotional factors facing
patients who are being asked to make decisions about their own treatment. ZikmundFisher
and coworkers suggest that physicians should consider asking patients to “reframe”
their decision by thinking about it as if it were a decision they were making for
someone else. The idea behind doing this would be to help the patient gain a better
understanding of the trade-offs they face.
Test Yourself 12.2
1. What is inductive reasoning, and how is it different from deductive reasoning?
2. How is inductive reasoning involved in the practice of science? How do inductive
and deductive reasoning work together in scientific research?
3. How is inductive reasoning involved in everyday experience?
4. How do the following cause errors in reasoning: availability heuristic; illusory correlations;
representativeness heuristic; confirmation bias?
5. How can failure to take into account base rates and small sample sizes cause errors
in reasoning?
6. What is the cross-cultural evidence regarding how U.S. and Itza participants use
typicality and diversity when making inferences about categories? How are the
behaviors of these two groups similar? Different? What is an explanation for the
differences?
7. What is the utility approach to decisions? What are some problems with the utility
approach? As you consider this, take into account mental simulations and the focusing
illusion.
8. How does the way a problem is stated and the need to justify decisions affect the
decisions people make?
9. How is the prefrontal cortex involved in problem solving and reasoning? Cite evidence
from both neuropsychology and brain imaging experiments to support your
answer.
10. What is neuroeconomics? Describe Sanfey and coworkers’ (2003) experiment, and
indicate what it adds to our understanding of decision making.
479Reasoning and Decision Making
11. How are people’s decisions about treatment options influenced by the person or
group for whom they are making the decision?
ChapterSummary
1. Reasoning is a cognitive process by which people start with information and come
to conclusions that go beyond that information. Deductive reasoning involves
syllogisms and can result in definite conclusions. Inductive reasoning is based on
evidence and results in conclusions that are probably true.
2. Categorical syllogisms have two premises and a conclusion that describe the relation
between two categories by using statements that begin with All, No, or Some.
3. A syllogism is valid if its conclusion follows logically from its premises. If the premises
of a valid syllogism are true, then the conclusion must be true. People are able
to correctly judge the validity of Aristotle’s “perfect” syllogism, but make errors in
all other forms of categorical syllogisms. Two reasons for errors are the atmosphere
effect and the belief bias.
4. One way to determine whether a categorical syllogism is valid or invalid is to use
diagrams. Another method is to create mental models representing the premises.
5. Conditional syllogisms have two premises and a conclusion, like categorical
syllogisms, but the first premise has the form “If . . . then. . . .” The four basic types
of conditional syllogism are (a) affirming the antecedent and (b) denying the consequent
(both valid), and (c) affirming the consequent and (d) denying the antecedent
(both invalid).
6. The Wason four-card problem has been used to study how people think when evaluating
conditional syllogisms. People make errors in the abstract version, but perform
better for versions of the problem that are restated in real-world terms, such as
the “drinking age” and “postal” versions. The key to solving the problem is to apply
the falsification principle.
7. Based on experiments using different versions of the Wason problem, a number
of mechanisms have been proposed to explain people’s performance. These mechanisms
include using permission schemas, and the evolutionary approach, which
explains performance in terms of social-exchange theory. Many experiments have
provided evidence for and against these explanations, leaving the controversy about
how to explain the Wason problem still unresolved.
8. In inductive reasoning, conclusions follow not from logically constructed syllogisms,
but from evidence. Conclusions are suggested in inductive reasoning, with varying
degrees of certainty. The strength of an inductive argument depends on the representativeness,
number, and quality of observations on which the argument is based.
9. Inductive reasoning is one of the basic mechanisms for developing scientific theories
and evaluating scientific evidence. It also plays a major role in everyday life
because we often make predictions about what we think will happen based on our
observations about what has happened in the past.
480 Chapter 12
10. The availability heuristic states that events that are more easily remembered are
judged as being more probable than events that are less easily remembered. This
heuristic can sometimes lead to correct judgments, and sometimes not. Errors due
to the availability heuristic have been demonstrated by having people estimate the
relative prevalence of various causes of death.
11. Illusory correlations and stereotypes, which can lead to incorrect conclusions about
relationships between things, are related to the availability heuristic, because they
draw attention to specific relationships and therefore make them more “available.”
12. The representativeness heuristic is based on the idea that people often make judgments
based on how much one event resembles another event. Errors due to this
heuristic have been demonstrated by asking participants to judge a person’s occupation
based on descriptive information. Errors occur when the representativeness
heuristic leads people to ignore base-rate information. In other situations, judgment
errors occur when people ignore the conjunction rule and the law of large
numbers.
13. The confirmation bias is the tendency to selectively look for information that conforms
to a hypothesis and to overlook information that argues against it. Operation
of this bias was demonstrated by Wason’s number-sequence task. This bias also operates
in real life when people’s attitudes influence the way they evaluate evidence.
14. Cultural differences in inductive reasoning have been demonstrated by comparing
how Itza and American participants apply the similarity-coverage model and
the associated typicality and diversity principles to their conceptions of different
categories. Experiments have demonstrated that these groups apply the typicality
principle similarly but differ in their application of the diversity principle.
15. The utility approach to decision making states that people are basically rational,
so when they have all of the relevant information, they will make a decision that
results in outcomes that are in their best interest. Evidence that people do not always
act in accordance with this approach includes failure to act in a way to maximize
monetary value, errors caused by application of mental simulations, and errors
caused by the focusing illusion.
16. Decisions can depend on how choices are presented, or framed. For example, when
a choice is framed in terms of gains, people use a risk-aversion strategy, and when a
choice is framed in terms of losses, people use a risk-taking strategy. Decision making
is also influenced by people’s tendency to want to justify their decision and state
a rationale for the decision.
17. The prefrontal cortex (PFC) is one of the major areas of the brain involved in
thinking. Damage to the PFC can cause perseveration and poor planning ability,
which results in poor performance on everyday tasks, problems such as the Tower
of Hanoi and water-jug problem, and some basic problems in reasoning.
18. Neuroeconomics studies decision making by combining approaches from psychology,
neuroscience, and economics. The results of a neuroeconomics experiment using
the ultimatum game have shown that people’s emotions can interfere with their
ability to make rational decisions. Brain imaging indicates that the anterior insula is
481Reasoning and Decision Making
associated with the emotions that occur during the ultimatum game and also suggests
that the PFC may be involved in the cognitive demands of the task.
19. An experiment that involved asking people to make a risky decision about being
vaccinated against a deadly disease has shown that people are more likely to recommend
that others receive the vaccination than they are to choose to receive the vaccination
themselves. This result has implications about how physicians talk about
treatment options with their patients.
Think About It
1. Astrology is popular with many people because they perceive a close connection between
astrological predictions and events in their lives. Explain factors that might
lead to this perception, even if a close connection does not, in fact, exist.
2. Think about a decision you have made recently. It can be a minor one, such as deciding
which restaurant to go to on Saturday evening, or a more major one, such as
picking an apartment or deciding which college to attend. Analyze this decision,
taking into account the processes you went through to arrive at it, and how you
justified it in your mind as being a good decision.
3. Create deductive syllogisms and inductive arguments that apply to the decision you
analyzed in the previous question.
4. Johanna has a reputation for being extremely good at justifying her behavior by a
process that is often called “rationalization.” For example, she justifies the fact that
she eats anything she wants by saying “Ten years ago this food was supposed to be
bad for you, and now they are saying it may even have some beneficial effects, so
what’s the point of listening to the so-called health experts?” or “That movie actor
who was really into red meat lived to be 95.” Analyze Johanna’s arguments by stating
them as inductive or deductive arguments or, better yet, do that for one of your
own rationalizations.
5. From watching the news or reading the paper, what can you conclude about how
the availability heuristic can influence our conceptions of the nature of the lives
of different groups of people (for example, movie stars, rich people, various racial,
ethnic, or cultural groups) and how accurate these conceptions might actually be?
6. Describe cases in which you made a poor decision because your judgment was
clouded by emotion or some other factor.
If You Want to Know More
1. Using diagrams to determine the validity of syllogisms. There are a number
of ways to use diagrams to determine whether syllogisms are valid. One method
uses Venn diagrams; another uses Euler circles.
482 Chapter 12
Edwards, A. W. F. (2004). Cogwheels of the mind: The story of Venn diagrams. Baltimore:
Johns Hopkins University Press.
Shin, S-J. (1994). The logical status of diagrams. Cambridge, UK: Cambridge University
Press.
2. Culture and cognition. We have seen that culture can affect inductive reasoning
(see page 463 in this text). Other studies in which reasoning has been tested in different
cultural groups (and usually compared with a sample of participants from the United
States) have investigated how people from different cultures solve syllogisms, reason
inductively, and solve math problems.
Dehaene, S., Izard, V., Pica, P., & Spelke, E. (2006). Core knowledge of geometry in an
Amazonian indigene group. Science, 311, 381–384.
Nisbett, R. E. (2003). The geography of thought. New York: Free Press.
Scribner, S. (1977). Modes of thinking and ways of speaking: Culture and logic reconsidered.
In P. N. Johnson-Laird & P. C. Wason (Eds.), Thinking: Readings in cognitive
science (pp. 483–500). Cambridge: Cambridge University Press.
Tang, Y., Zhang, W., Chen, K., Feng, S., Ji, Y., Shen, J., Reiman, E. M., & Liu, Y.
(2006). Arithmetic processing in the brain shaped by cultures. Proceedings of the National
Academy of Sciences, 103, 10775–10780.
3. Making decisions rapidly. The deliberation-without-attention effect states that
under certain conditions, people make better decisions when they act quickly than when
they consciously deliberate over the decisions.
Dijksterhuis, A., Bos, M. W., Nordgren, L. F., & van Baaren, R. B. (2006). On making
the right choice: The deliberation-without-attention effect. Science, 311, 1005–1007.
4. Reasoning and the law. Juries are asked to come to a conclusion by evaluating
evidence. New research indicates that a person’s beliefs can affect his or her decision
making. Also, the brain’s response depends on whether the evidence is consistent or
inconsistent with the person’s beliefs.
Fugelsang, J. A., & Dunbar, K. N. (2004). A cognitive neuroscience framework for understanding
causal reasoning and the law. Philosophical Transactions of the Royal Society
of London, B: Biological Sciences, 359, 1749–1754.
5. Regret and decision making. The results of the ultimatum game experiments
show that human decision making can be influenced by emotions. Other experiments
have shown that people’s decision making is influenced by a desire to avoid the regret
they would experience if they made the wrong decision.
Coricelli, G., Critchley, H. D., Joffily, M., O’Doherty, J. P., Sirigu, A., & Dolan, R. J.
(2005). Regret and its avoidance: A neuroimaging study of choice behavior. Nature
Neuroscience, 8, 1255–1262.
483Reasoning and Decision Making
6. Another view of rationality. We described the idea that people’s reasoning and
decision making can be negatively affected by bias and the use of heuristics. Some researchers
have proposed another approach that sees people as behaving more rationally
than this view gives them credit for.
Chase, V. M., Hertwig, R., & Gigerenzer, G. (1998). Views of rationality. Trends in Cognitive
Sciences, 2, 206–214.
7. Embodied cognition. Some cognitive psychologists propose that to understand
cognition, we need to take into account how a person’s body interacts with the environment.
This idea has implications for understanding problem solving, reasoning, and
language.
Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin and Review, 9,
625–636.
8. Neurons that respond to abstract rules. There are neurons in the monkey
PFC that respond to abstract rules. This adds to the evidence that the PFC is important
for problem solving.
Wallis, J. D., Anderson, K. C., & Miller, E. K. (2001). Single neurons in prefrontal cortex
encode abstract rules. Nature, 411, 953–956.
Key Terms
Affirming the antecedent, 444
Affirming the consequent, 444
Antecedent, 443
Atmosphere effect, 439
Availability heuristic, 456
Base rate, 460
Belief bias, 439
Categorical syllogism, 437
Conclusion, 437
Conditional syllogism, 443
Confirmation bias, 462
Conjunction rule, 461
Consequent, 443
Decisions, 437
Deductive reasoning, 436
Denying the antecedent, 444
Denying the consequent, 444
Diversity principle, 464
Economic utility theory, 466
Evolutionary perspective on cognition, 450
Falsification principle, 446
Focusing illusion, 467
Framing effect, 471
Illusory correlation, 458
Inductive reasoning, 436
Law of large numbers, 462
Mental model, 441
Mental simulation, 467
Neuroeconomics, 475
Omission bias, 477
Opt-in procedure, 469
Opt-out procedure, 469
Permission schema, 448
Perseveration, 473
Pragmatic reasoning schema, 448
Premise, 437
Reasoning, 435
Representativeness heuristic, 459
Risk-aversion strategy, 470
Risk-taking strategy, 470
484 Chapter 12
To experience these experiments for yourself, go to http://coglab.wadsworth.com. Be
sure to read each experiment’s setup instructions before you go to the experiment itself.
Otherwise, you won’t know which keys to press.
Primary Labs
Wason selection task Two versions of the Wason four-card problem (p. 445).
Typical reasoning How the representativeness heuristic can lead to errors of judgment
(p. 459).
Risky decisions How decision making is influenced by framing effects (p. 469).
Decision making An experiment that demonstrates how decisions can be affected by
the context within which the decision is made (p. 469).
Related Lab
Monty Hall A simulation of the Monty Hall three-door problem, which involves an
understanding of probability.
The cue ball is to the left of your
line of sight.
B
RCue G
■ Figure 12.21 Result of visualization of pool balls (see page 441).
Similarity-coverage model, 464
Social-exchange theory, 450
Stereotype, 459
Syllogism, 436
Typicality principle, 464
Ultimatum game, 475
Utility, 466
Valid, 438
Wason four-card problem, 445