The Value of Cognition § Humans owe their success more to thinking abilities rather than to physical strength or speed. § Homo sapiens • From the Latin for man and wise. § Our mental abilities make us highly adaptable.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Value of Cognition
§ Humans owe their success more to thinking abilities rather than to physical strength or speed.
§ Homo sapiens • From the Latin for man and wise.
§ Our mental abilities make us highly adaptable.
§ “Cognition” is derived from the Latin cogito, which means “to think.”
§ Thinking allows us to…
• Manipulate information internally
• Construct models of the world
• Plan our interactions with that world
• Regulate ourselves to meet our goals.
§ Let’s begin thinking about cognition by considering
a basic building block of thought – mental images.
Cognition
Mental Images
} “Mental image” = a representation of sensory experience that is stored in memory & can be
retrieved for use later.
§ E.g., “Picture” the letters of your name.
} We treat mental images much like a real object.
§ E.g., Turn mental images around in our minds, zoom in or out, and identify their features.
Mental Images
} Children are particularly likely to use visual images in their thinking
} Adults tend to rely more on verbal
representations than images.
§ One explanation:
• Language begins to organize thinking during
childhood.
• Use of language overwrites with the ability to directly
access visual images.
§ Not inaccessible, however – requires more effortful
processing.
Concepts
} Mental representations would be useless to us unless we imposed some type of organization.
§ We extract organizing ideas known as conceptsfrom experiences.
• In essence, concepts organize mental images.
} Concepts can be formal or natural
§ Formal concepts defined learned rules that certain categories of things.
§ Natural concepts that develop naturally as we experience the world.
How Are Concepts Defined?
} We construct rules that dictate natural concepts as we learn about our world.
• As a result, the boundaries defining natural concept
categories are often “fuzzy.”
§ Major flaw in this system.
• No matter how careful your definitions of concepts
are, somebody will be able to think of an exception!
§ “Cats are four-legged animals.”
How Are Concepts Defined?
1 2 34
§ “Cats are four-legged animals.”
How Are Concepts Defined?
1 2 34
1
2 3
§ “Cats are four-legged animals.”
How Are Concepts Defined?
1 2 34
1
2 3
Tripod Cat Falsifies Your
Concept!
§ “Cats are four-legged animals.”
How Are Concepts Defined?
1 2 34
1
2 3
Tripod Cat Falsifies Your
Concept!
Aren’t these two a fuzzy set?(That’s the worst joke of the
semester, if you are keeping track!)
How Are Concepts Defined?
} To resolve this dilemma, we can make our definition more flexible.
§ A concept could describe a group of instances that share overlapping features
• i.e., not a “checklist” of features that conform to
rules.
} This approach is similar to a feature detection model.§ There are several problems with this approach,
too.
• E.g., Some categories are quite clear (e.g., triangles), but
others do not have precise enough boundaries.
How Are Concepts Defined?
} Conjunctive Concepts: Defined by the presence of two or more features – “and”
} Disjunctive Concepts: Have at least one of
several possible features. “either/or”
} Relational Concepts: Based on how an object relates to something else, or how its features
relate to one another.
} Faulty Concepts: inaccurate concepts that lead to thinking errors. (e.g., social stereotypes are oversimplified concepts of groups).
Comparing Types of Concepts
} An alternate approach to thinking about concepts is to consider some type of
“standard” or “ideal model”
§ Prototype that represents your entire category.
• Results from an averaging of all the
members of a category
• May not even resemble any real instance!
} When thinking about a category, we
might also retrieve a specific instance of a concept, or an exemplar.
Comparing Types of Concepts
Comparing Types of Concepts
Comparing Types of Concepts
} Representing concepts in terms of exemplars has advantages over prototypes.
§ E.g., Exemplars provide a better way of thinking about the variability of a category.
• Prototypical averages don’t provide information
about the range of features that can be found in a
category!
Comparing Types of Concepts
Comparing Types of Concepts
Comparing Types of Concepts
} How do we organize concepts?
} We tend to organize our knowledge into three levels of categorization:
§ Superordinate
§ Basic
§ Subordinate
Organizing Concepts
} The superordinate category contains concepts that are broad and general.
• E.g., “Fruit”
} The intermediate basic level category is what we
typically use to think about our world.
• E.g., “Oranges”
} Concepts at the subordinate level are less
general and more specific than those at the basic level.
• E.g., “Valencia Oranges”
Organizing Concepts
Concepts as Theories
} Theories - Sets of facts and relationships between facts that can be used to explain and
predict phenomena.
} Concepts develop similarly to theories:
§ Guide our thinking
§ Continually tested for accuracy against new, incoming information.
§ Do not exist in isolation.
• Can be viewed as part of a vast, interconnected
network of memories.
Concepts as Theories
} “Concepts as theories”
provides insight into the
problem of judging category
membership
§ Prototypes and exemplars
provide a useful starting
place for judging category
membership.
§ We test our theory that the
new item fits the category by
comparing it to the
prototypes and exemplars of
a concept.
Schemas
} Concepts are embedded in a rich, complex set of beliefs and expectations and personal
experience known as schema.
§ Representations of a concept stored in memory.
§ Used to guide behavior and interpret new situations.
} Scripts are schema that describes how a series
of actions should unfold.
§ E.g., What happens when you go to a restaurant.
• When a new situation deviates from our script, we
may be confused about how to behave!
Problem Solving
} Some types of problems lend themselves to precise, step-by-step rules for reaching a
particular solution
§ Such “algorithms” have the advantage of producing an accurate solution reliably.
} One such algorithm, utility theory, is widely used in economics.
Problem-Solving: Algorithms
} Utility theory: we compute the expected outcomes of our choices and select the best one.
§ May be useful when parameters are clear & reliably estimable within a reasonable range (i.e., well-structured problems)
• However, that we rarely make decisions by solving
equations!
§ Seemingly rational choices (i.e., holding utility constant)
can be overridden by framing» Recall: Hsee, Abelson, & Salovey (1991).
Problem-Solving: Algorithms
} Ill-structured problems are problems for which there is no known algorithm
§ Intuition (i.e., simply believing that something is true independent of any reasoning process)
§ Intuition is fallible!
• E.g., We may not even think of certain possible
solutions.
Problem-Solving: Algorithms
Problem-Solving: Heuristics
} Availability
} Representative
} Recognition
} Affect
§ The availability heuristic is used when people predict that events that are easy to think about will be more frequent.
• E.g., Shark attacks; Airplane crashes
§ Which is more common, being killed by a shark or by falling
airplane parts?
» 30 times the risk of being killed by falling airplane parts than
by sharks.
§ After 9/11 more Americans chose to drive rather than fly.
» The extra traffic led to an ~9% increase in automobile
fatalities in the 3 months following the attacks.
Problem-Solving: Heuristics
§ The representativeness heuristic leads people to estimate that stimuli which are similar to a prototype are more likely to fit the category than are stimuli which are different from the prototype.
• For example, is Thomas, who is short, slim, and loves
poetry, more likely to be an Ivy League classics
professor or a truck driver?
Problem-Solving: Heuristics
§ Heuristics don’t always lead to bad decisions!
• Quick, effective, and efficient decisions were a significant adaptive advantage for our ancestors!
§ E.g. The recognition heuristic predicts that people
will place a higher value on the more easily
recognized alternative.
§ E.g., The affect heuristic suggests that we use our
emotional responses to each choice to guide our
decisions.
» Described as a “gut” reaction.
Problem-Solving: Heuristics
Barriers to Problem Solving
} Framing – i.e., formulating the problem
§ Takes time and effort
} Functional fixedness
§ Tendency to think about a concept in its most typical form and no
others.
} Mental sets
§ Tendency to habitually use the methods of problem solving that
have worked for you in the past.
Intelligence
} This section covers:
§ Assessing intelligence
§ Conceptualizations of intelligence
History of Assessing Intelligence
} Alfred Binet
} Theodore Simon
} Mental age vs. chronological age
§ “Intelligence” -- enduring abilities that allow you to adapt to your environment and behave in goal-
directed ways.
§ Historically, developing a precise definition of intelligence has been difficult!
§ Several revisions to this definition over time.
§ Key Question: How do we measure intelligence?
History of Assessing Intelligence
§ Modern intelligence test credited to Alfred Binet(1857–1911).
§ 1904
• French government wants develop a means of
measuring the intelligence of French schoolchildren
• Government wanted identify children who would not
likely profit from traditional education.
• Alfred Binet & Théodore Simon appointed to task.
History of Assessing Intelligence
§ Binet saw intelligence as…
§ The capacity to find and maintain a purpose
§ Adopt a strategy to reach that purpose
§ And evaluate the strategy so it can be adjusted as necessary.
• i.e., intelligence = good problem solving.
§ Developed an intelligence test that assessed general cognitive abilities that aid in problem
solving.
§ E.g., attention, judgment, and reasoning skills.
History of Assessing Intelligence
§ 30 tasks that measured these skills
§ Arranged them in order of difficulty
• Easiest questions first => hardest questions last.
§ Observations:
§ “Brighter” students could answer more of the questions.
§ Older children tended to answer more questions correctly.
History of Assessing Intelligence
§ Younger children could sometimes answer correctly as many questions as the average child of
an older age.
§ E.g., Very smart 6-year-old might be able to answer as many questions as the average 10-year-old child could.
§ Binet began to quantify children’s intelligence in terms of mental age§ Age that reflects mental abilities in comparison to the
“average” child.
History of Assessing Intelligence
§ A mental age that exceeds one’s chronological age indicates above-average intelligence
§ A mental age that is below a child’s actual age
indicates a below-average level of intelligence.
§ The foundation for the IQ score
§ This test became the basis for modern intelligence tests.
History of Assessing Intelligence
Modern Forms of Assessing Intelligence
} Stanford-Binet
} Wechsler Intelligence
Scales
§ 1916
§ Lewis Terman -- American revision of the Binet and Simon test.
§ “Stanford Revision of the Binet-Simon Scale”
• i.e., the Stanford-Binet.
§ Standardized test
• A test that uses a standard set of questions,
procedures, and scoring methods for all test takers.
Modern Forms of Assessing Intelligence
§ To standardize the Stanford-Binet, Termandeveloped age-based norms.
§ Terman gave the test to a large number of people
§ Calculated the average test scores for people of different ages.
§ Such norms allowed Terman to establish mental age scores for people taking the Stanford-Binet.
Modern Forms of Assessing Intelligence
§ Terman popularized the use of an intelligence quotient, or IQ
§ IQ score is a person’s mental age divided by chronological age, then multiplied by 100.
• A person of average abilities has, by definition, an IQ
of 100.
§ i.e., A mental age equal to their actual age.
§ IQs over 100 indicate above-average intelligence.
§ IQs below 100 indicate below-average intelligence.
Modern Forms of Assessing Intelligence
§ Stanford-Binet has undergone four major revisions since 1916
§ Still in wide use today.
• Fifth Edition (SB5), was released in 2003.
§ David Wechsler (1896–1981) released a competing test in 1939 that greatly challenged the popularity of the Stanford-Binet.
§ In response to shortcomings he saw in the Stanford-Binet.
Modern Forms of Assessing Intelligence
§ Wechsler objected to the fact that the Stanford-Binet test tried to sum up intelligence in a single
score.
§ Can’t adequately express something as complex as intelligence in one summary score.
§ Also objected to the use of the mental age concept for adults.
• Would you necessarily expect a 40-year-old to
correctly answer more questions than a 35-year-old?
§ Adults do not change as much from year to year as children
do.
» Mental age has little significance in adulthood!
Modern Forms of Assessing Intelligence
§ Wechsler’s test yields scores on individual subscales that measure different mental abilities.
§ Wechsler’s tests compare a participant’s performance to the average person’s performance to determine IQ.
• Standardized tests
• Devised so that an average person’s performance on
the test results in an IQ of 100.
Modern Forms of Assessing Intelligence
§ Above average IQ scores are above 100, and below average are given IQ scores below 100.
§ Most people can expect to score near this average IQ, somewhere in the range of 85–115
§ Three separate Wechsler intelligence tests.
§ The Wechsler Preschool and Primary Scale of Intelligence (WPPSI-IV): Children ages 21/2 to 7.
§ The Wechsler Intelligence Scale for Children (WISC-IV); Children ages 6–16.
§ The Wechsler Adult Intelligence Scale (WAIS-IV): People ages 16–90.
Modern Forms of Assessing Intelligence
What Makes a Good Intelligence Test?
} Reliability
} Validity
§ Reliability -- refers to the degree to which the test yields consistent measurements over time.
§ Although intelligence can change over time, it usually does so very slowly.
§ In general, if you are intelligent today, you will be intelligent 6 months from now.
• So, if we use a test to measure your IQ today and then
again in 6 months, the scores should be comparable.
What Makes a Good Intelligence Test
§ Validity -- the degree to which the test measures what it was designed to measure.
§ In the case of an intelligence test, one must show that the test actually measures intelligence!
• For example, do scores on the test reliably predict
future behavior?
§ If we expect that intelligence is related to doing well in school,
then scores on a valid IQ test should predict who does well in
school and who does not.
What Makes a Good Intelligence Test
Conceptualizations of Intelligence
} Single factor
} Collection of abilities
} Multiple intelligences
} Triarchic intelligence
} Emotional intelligence
§ Test scores of separate mental abilities tend to correlate.
§ Charles Spearman argued that because of this, there must be one general level of intelligence that underlies these separate mental abilities.
• “G” for general mental ability.
§ By the 1930s, some theorists were beginning to
challenge the idea of a single intelligence.
§ Psychologists proposed theories that described intelligence as a set of abilities rather than a single trait.
The Nature of Intelligence
§ Thurstone argued that intelligence was made up of seven distinct mental abilities:
§ reasoning, associative memory, spatial visualization, numerical fluency, verbal comprehension, perceptual speed, and word fluency.
• Others would eventually propose as many as 120
different factors underlying intelligence!
The Nature of Intelligence
§ In the 1960s, Raymond Cattell (1963) revived the idea of “G”.
§ “G” does exist, but in two different forms:
§ Crystallized intelligence
• Our accumulation of knowledge.
§ Fluid intelligence.
• Speed and efficiency with which we learn new
information and solve problems.
The Nature of Intelligence
§ Good and bad news...
§ Crystallized intelligence can continue to grow well into late adulthood
§ Fluid intelligence tends to decrease across adulthood.
§ The degree to which we retain these abilities
throughout life is affected by numerous factors
§ Environment and Physical Well-Being
The Nature of Intelligence
The Nature of Intelligence
The Nature of Intelligence (cont’d.)
§ Educational Assessment
§ SAT, ACT, GRE
§ Pre-Employment Testing
Uses of Intelligence Testing
Related Documents
