Reasoning and Judgment

The process of complex thinking involves the three processes

of reasoning, judgment and decision making
Reasoning involves evaluation of a conclusion based
solely on given information
In Judgment reasoning is applied on the given information to
arrive at a conclusion
Decision making evaluation of the given information is done to
arrive at a judgment and based on the judgment a choice
among several possible alternatives is achieved.
The focus on Errors
The emphasis in research on reasoning, judgment and
decision making has been on mistakes people make. Daniel
Kahneman (1991) believes errors provide us with
informativeness. The conditions under which our thinking fails
us reveal important aspects of cognitive processing and inform
us how the process of reasoning, judgment and decision
making process works
How exactly do we know that a given chain of reasoning,
judgment and decision making is in error?

one approach normative approach describes how we ought to

think in a given situation, while a second approach descriptive
approach describes how we actually think.
Bounded Rationality
Adherence to or deviance from rational thoughts and
behavior depends on a variety of factors (an important one is
how we define rational). Baron (1999) believes rationality in
not necessarily the same as accuracy and that irrationality is
not the same as error. Rationality involves choosing the
methods that help us attain our goals. We can reason well but
still have a decision work out badly; conversely we can
reason badly yet still luck into a good outcome. The simple
notion that there are limits to our powers of reason is termed
bounded rationality.
Dual process views
According to this view the human thinker operates in
one of the two modes depending on the particular nature of
the situation. In heuristic mode (system 1), the process used
for thinking operate quickly and without much deliberation
(automatic) while in analytical mode (system 2) the processes
are relatively slow, deliberate and controlled. The analytical
mode is more cognitively demanding than the heuristic mode
in that it demands more working memory
 Reasoning
Deductive reasoning – involves determining if a specific
conclusion is valid based on general principle or assertions
(premises). Deductive reasoning problems involve a large
degree of constraint and the conclusion is easily assessed
using the algorithmic approach. Two primary types from Evans
(2002) are discussed

Syllogistic Reasoning
All students are bright
All bright people complete assigned work on time
Therefore, all students complete assigned work on time
 1) Syllogisms consists of two premises and a conclusion.

2) The premises & conclusions may begin with a

universal quantifier (all) or a particular quantifier (some).

3) Also the terms within a syllogism may be stated

positively (“All A are B”) or negatively (“All A are not B”).

Syllogism are either valid or not valid – that is the

conclusions either does or does not hold, given the premises
Valid arguments imply that the conclusions does follow from the
premises. However it says nothing about whether the premises
themselves are true. The truth value of an argument depends
on both the validity of the argument form and the truth of the
premises. Consider

All professor are comedians

All comedians are funny
Therefore, all professors are funny

All A are B
All C are B
Therefore, All A are C
 No oranges are apples
No lemons are oranges
Therefore, no apples are lemons

Are these conclusions valid based on their respective

premises? The reasoning errors are caused by
a) Atmosphere effects – according to this explanation the
quantifiers used in the premise combine to form an
“atmosphere” within which the validity of the conclusion is
assessed. For instance the premises in syllogism 1 create
 a “positive universal atmosphere”. This produces an
erroneous tendency to claim that the universal and positive
conclusion are valid.
b) Belief bias – our beliefs about truth interfere with our ability
to assess argument validity.
All intelligent beings are Simpsons fans
All dolphins are intelligent beings
Therefore, all dolphins are Simpsons fans
The tendency to allow belief to interfere with the evaluations of
conclusions in syllogistic arguments has been termed belief bias
All smart people are reasonable
All Democrats are smart people
Therefore, all Democrats are reasonable
Conditional Reasoning
the second form of deductive reasoning is called
conditional reasoning (or if –then reasoning) and involves
evaluating whether a particular conclusion is valid given that
certain conditions (premises) hold. Consider
1. if someone likes Winnie-the-Pooh, they are a
sensitive person.
2. Mary likes Winnie-the-Pooh.
3. Therefore, Mary is a sensitive person
Another version of the reasoning problem is
1. if someone likes Winnie-the-Pooh, they’re a sensitive
person
2. Mary is a sensitive person
3. Therefore, Mary likes Winnie-the-Pooh.
Conditional reasoning conclusions can be evaluated quite
easily if one applies a set of logical rules.
Conditional statement

if a person likes Winnie-the-Pooh, then they’re a sensitive

person
(Antecedent) (Consequent)
Four conditional scenarios
 Affirm Deny

Antecedent Mary like Winnie-the-Pooh Mary does not like Winnie-

Therefore, Mary is a the Pooh
sensitive person (Modus Therefore, Mary is not a
Ponens) sensitive person

Consequent Mary is a sensitive person. Mary is not a sensitive

Therefore, Mary likes person.
Winnie-the-Pooh Therefore, Mary does not like
Winnie-the-Pooh (Modus
Tollens)
People run into fair amounts of difficulty when judging the
validity of conclusions derived from if –then statements. One
tendency people have is to interpret the initial conditional
statement as bi-conditional- thinking that “if p, then q” also
means “if q, then p”
Wason’s Selection task
the classic version of the task is requires the reasoner to
decide which of the four cards needs to be turned over in order
to determine whether the following if-then-statement holds:
if a card has a vowel on one side, then it must have an
even number on the opposite side
The selection tendencies of WST reveal confirmatory bias –
which refers to our tendency to seek out or notice evidence
that is consistent with a particular hypothesis rather than
evidence that would be inconsistent with the hypothesis.
Rules or Models of deductive reasoning
Explanations' of how we reason deductively generally fall into two
main camps
1) strict or rule based account (Rips, 1994) – which contends that
people possess the representational equivalent of logic rules. These
rules are applied to the premises to determine if the conclusion is
valid
2) mental model view (Johnson-Laird, 2002) – which believes that
we first form a mental model based on the information in the
premises and out own previous experiences. Next we search for a
mental model in which the premises would be true but the stated
conclusion would be false. If successful we deem the conclusion
invalid; if we don’t we deem the conclusion valid.
Inductive Reasoning
In inductive reasoning we reason from specific pieces of
data or information towards a general conclusion. Unlike
deductive reasoning where conclusions are labeled as
valid/invalid with absolute certainty, inductive reasoning leads
to uncertain conclusions that vary in their strength
Professor X gets upset when asked if she’ll issue paper extension
Professor Y won’t accept late papers
Professor Z takes 20% off each day a paper is late
Bisanz, Bisanz & Korpan (1994) describe some characteristics
that seem to typify inductive reasoning
1) The product of inductive reasoning is not necessarily
correct. Inductive arguments are evaluated in terms of their
strength rather than in terms of their validity.
2) Ripps (1990) points out that with inductive reasoning
there is a need for constraint on the conclusion to be reached
Rules or Instances?
what mental structures & processes underlie inductive
reasoning?
Researchers disagree on whether induction is based on
formal, rule-driven processes or on more context-bound,
experience based heuristic processing. The rule based view –
termed as strict/syntactic view – states that inductive reasoning
involves special processes and representations that operate in
the abstract, outside of any real life context. The experience-
based view – termed loose view – contends that inductive
reasoning involves updating the strengths of one’s belief based
on the recall of specific instances
The Omnipresence of Inductive Reasoning
Inductive reasoning is in some part involved with most of the
cognitive processes. Two examples are
a) Inductive reasoning in Categorization – inductive
reasoning provides another view through which the
phenomenon known as the typicality effect can be viewed.
1. Robins are susceptible to disease A
Therefore, all birds are susceptible to disease A
2. Turkeys are susceptible to disease B
Therefore, all birds are susceptible to disease A
subjects rated argument 1 as more likely to be true because
robins are seen as more typical birds than turkeys
Another interesting phenomenon observed in inductive
reasoning about categories might be termed as diversity
effect. Which of the following arg. is strong
1. Robins are susceptible to disease Y
Sparrows are susceptible to disease Y
Therefore, all birds are susceptible to disease Y
2. Cardinals are susceptible to disease Z
Turkeys are susceptible to disease Z
Therefore, all birds are susceptible to disease Z
In this case people rated argument 2 as stronger because
cardinals and turkeys represent a more diverse set of birds
relative to robins and sparrows
 (b) Inductive reasoning and problem solving
another set of cognitive processes that depends critically
on inductive reasoning is problem solving – more specifically
solving problems by analogy.

Judgment
Inductive reasoning involves arriving at general
conclusions based on specific pieces of what might be called
“data”. Judgment is an extension of inductive reasoning. Hastie
& Dawes (2001) define –
judgment is the human ability to infer, estimate & predict
the character of unknown events.
Judgment is the process of making educated guesses, based
on limited information along with our previous knowledge,
expectations and beliefs (stereotypes)

Basing judgment on memory: The availability heuristic

the availability heuristic indicates that we base our
estimates of likelihood, or probability, on the ease with which we
can think of examples. The availability heuristic is dependent on
two main sub processes
a) biased encoding – leads to overrepresentation of certain facts in
memory. This in turn makes bias retrieval from memory as the
information stored is memory is biased. (e.g., media overestimation)
b) biased retrieval – availability can lead us astray if the sampling
process itself is biased. (e.g., Try the following and state whether
there are more of number 1 or number 2)
1. six letter words that have the letter n as the fifth letter
2. words that fit the pattern _ _ _ i n g.
c) illusory correlations – when one notices primarily coincidences,
two events will seem to be linked even when they’re not. This
perception is called illusory correlation (sports illustrated jinx)
A recognition Heuristic – reasoning is adaptive, we simply
cannot consider all of the data, nor do we have access to it.
The lack of data can be informative!. Recognition heuristic is
often used when we’re faced with two alternatives – one that’s
recognizable and one that’s not.
The Representative Heuristic: Basing judgment on
similarity
When trying to place a person in a particular category our
judgment rely on representational heuristic – the degree to
which the object represents our basic idea of that object
Ignoring base rates
people mostly commit the base rate fallacy – ignore thee rate of
occurrence of a particular category in the population or sample (i.e.,
how often a certain event tends to occur) – there by getting biased
by similarities. For example consider the classic demonstration by
Kahneman & Tversky (1973).
Subjects were given the following instructions
a panel of psychologists have interviewed and administered
personality tests to 30 engineers and 70 lawyers, all successful in
their field. On the basis of this information, thumbnail descriptions for
each of these individuals have been written. For each description,
please indicate the probability that the person described is an
engineer from 1 to 100.
Subjects were then given the following description
jack is a 45-year-old man. He is married and has 4
children. He is generally conservative, careful and
ambitious. He shows no interest in political and social
issues and spends most of his times on his many hobbies,
which include home carpentry, sailing and mathematical
puzzles.
Subjects were required to rate the probability that jack was
an engineer.
when not given the profile the probability that a
randomly drawn name is an engineer was 30/100 or 30%,
however with profile it was 50/100 or 50%.
Use of representative heuristic and the concomitant tendency to
ignore base rate may relate to the use of the controversial
practice known as racial profiling.
The conjunction fallacy
the conjunction fallacy is another cause of bias which is
caused by stereotyping. In a classic study Tversky & Kahneman
(1983) demonstrated this fallacy at work. They presented the
following problems to subjects
Linda is 31 years old; she’s single, outspoken and very
bright. She majored in philosophy. As a student, she was deeply
consumed with issues of discrimination and social justice, and
also participated in anti-nuclear demonstrations.
The subjects were asked to decide whether it was more likely that
she was (1) a bank teller (2) a bank teller who was active in the
feminist movement.
if probability of 1 was 0.5 and that of 2 was 0.5 then since
their combined probability (since 1 includes 2) is 0.5 x 0.5 = 0.25
Misperception of event clusters
when a given event has two different ways of working out,
such as a coin flip, people tend to misconstrue what a random
sequence should look like (i.e.,) they tend to underestimate the
number of clusters of like event, that would occur in a truly
random sequence.
HTHTTHTHHTHT HHHHHHTTTTTT
Examples of misperception of event clusters
a) The hot hand – the tendency to misperceive event
clusters as indicating non-randomness may underlie
what sports fans terms as hot hand.
b) The Gamblers fallacy – refers to the belief that after a
run of bad luck (or a run of a certain type of outcome), a
change is “due” to occur
The Anchoring-and-Adjustment Heuristic
In many cases of judgment people start with an idea or
standard in mind. This initial estimate or first impression tends to
make us overly biased towards it. The heuristic involved in
these judgment is termed anchoring and adjustment. A good
example of this heuristic at work is the spotlight effect.
spotlight effect – refers to our tendency to believe that
others notice our actions and appearance more than they
actually do – i.e., we believe that the “social spotlight” shines
more brightly on us than it actually does.
Biased Evaluation of Our Judgments
Biases in judgment can also arise from the fact the at
times we’re not good at estimating how much we know or
when we knew it. A couple of biases of this sort are
a) Hindsight Bias: people often seem to be sure after
something has occurred that they knew things would work
out just that way. This tendency is termed as hindsight bias
(i-knew-it-all-along-effect). [e.g., civil suits]
b) Miscalibration of Confidence: the fact that we
overestimate the extent to which we knew something we
going to happen demonstrates an insensitivity to what we
knew and when we knew it.