Cognitive Psychology, 2nd Ed.

1
Cognitive Psychology, 2nd Ed.

• Chapter 9
• Learning Concepts and Skills

2
Concept Learning

• Understanding space-time is fundamental. Object
  permanance allows an infant to represent objects
  are permanent entities that do not vanish when
  out of sight.
• 4-5 month olds make the AB error in searching for
  a missing object. By 9 months the concept is
  grasped. But even at 5 months preferential
  viewing of a novel number of objects suggests the
  beginnings of object permanance and numerosity.

3
(No Transcript)
4
Concept Learning

• Basic concepts constitute the most informative
  level of object categorization. Its
  characteristic features are common and correlated
  (e.g., bird). Subordinate concepts are even more
  so, but carve the world at too fine a grain (e.g.
  crow). Superordinate concepts are too general
  (animal).
• Frequency of exposure rather than hierarchical
  level appears to determine age of acquisition.
  Infants learn dog-cat-bird contrasts because of
  frequent examples.

5
Learning Processes

• The frequency of occurrence of stimulus features
  is automatically accumulated, providing a basis
  for identifying defining features and typicality
  effects.
• Learners also actively test hypotheses, requiring
  controlled processing in working memory, to
  identify defining features.

6
(No Transcript)
7
(No Transcript)
8
Prototype Acquisition

• Abstraction of a central tendency prototype may
  involve processes that calculate either mean,
  median, or modal values. Instead a frequency
  count and perceptual error may account for
  apparent variations. The mode may only look like
  the mean or median.
• Nonanalytic strategy of storing specific
  examplars is an alternative to abstraction.
  Cortical dissociation in neuroimaging plus
  behavioral data suggest exemplar storage is
  insufficient by itself to explain abstraction.

9
(No Transcript)
10
(No Transcript)
11
Implicit Learning

• Refers to the unconscious acquisition of complex
  rules that cannot be verbalized. Implicit tasks
  involve many stimulus dimensions related by
  complex rules, not simple conjuctions and
  disjunctions.
• Artificial, finite state grammars reflect the
  complexity of natural language grammars.
  Learners can categorize new strings but cannot
  articulate the underlying rules. Implicit
  learning or examplar storage are contrasting
  explanations.

12
(No Transcript)
13
(No Transcript)
14
Implicit vs.Explicit Learning

• Connectionist (frequencies) or symbolic (rule).
• U-Mode-unselective, passive, unreportable.

• Symbolic rule or specific exemplars.
• S-Mode-selective, effortful, and reportable.

15
Acquiring Expertise

• Skills involve perceptual, motor, and cognitive
  functioning. Contrast running versus reading,
  for example.
• Skills require procedural nondeclarative memory,
  but declarative memory is further involved in
  learning about a domain (e.g., contrast a pro
  football player with a surgeon).

16
Expert-Novice Differences

• Refer to differences in mental representations
  and cognitive processes associated with
  domain-specific expertise.
• Comparisons of average person with world-class
  performer in chess, mathematics, music, dance
  tennis, skating, medical diagnosis and more have
  shed light on skill learning.

17
(No Transcript)
18
Skill Acquisition

• Early, intermediate, and final stages of
  acquisition moves from declarative, controlled
  learning to procedural, automatic performance.
• Principal of procedural reinstatement states that
  a skill is well-retained for long periods of time
  when the procedures used to acquire it are
  reinstated precisely at test. Skills can be
  immune from forgetting, but transfer to new
  conditions negligible.
• Deliberate practice for 10 years minimum needed
  to attain world class performance.

19
Expert-Novice Differences

• Folk theories of novices cause reasoning errors
  in physics problems. Experts unaffected by
  problem context.
• Metacognitive control--experts reflect on problem
  first before calculating.
• Experts mnemonically encode information and
  develop sophisticated retrieval structures for
  rapid access to information in long-term memory
  (sometimes called long-term working memory).
• Speed increases and errors decrease with practice
  according to a power law.

20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)