I.  Representational hypothesis

A. young children’s concept are concrete, perceptual, holistic, thematic, and global

B. older children’s and adults’ concepts are abstract, conceptual, analytic, taxonomic, and specific

C. how accurately do such distinctions capture young children’s conceptual understanding?

D. three possibilities for conceptual representations

1. defining-features (dictionary)
2. probabilistic features (encyclopedia)
3. theory-based representations (textbook)

II.  Defining Feature Representations

A.  early research

1.  thematic vs. taxonomic categories (Piaget)
2.  thematic --> chain --> true concepts (Vygotsky)

B.  Evidence for sensitivity to defining features representations

1.  Bauer & Mandler (1989) with 1 year-olds in a pointing task

a. taxonomic matches were made 85% of the time

2.  Smith (1984) with 2-4 year-olds in an imitation task

a. 2 exps demonstrated a concept (e.g. red) by holding up different size and colors of multidimensional stimuli (e.g., flowers)
b. by age 3, children consistently showed conceptual knowledge

C.  Preference vs. ability (Smiley & Brown, 1979)

1. pre-, 1st, 5th, college, and elders grouped pictures (e.g. sheep-goat-wool)
2. 5th graders and college students showed taxonomic relations, others thematic
3. however, 1st and elders were able to specify logic for both methods
4. preference, not necessarily a change in coding

D. Role of specific knowledge (Keil & Batterman, 1984)

1. 5 and 9 year olds could use defining features for better known concepts (e.g., robbers), but both had problems for lesser known ones (e.g., taxi)

III.  Probabilistic Features Representations

A. based on family resemblance or prototypes

1. 4 main ideas

a. cue validities, basic level categories, correlations among features, prototypes

B.  Evidence

1. basic level categories (e.g., Columbo et al., 1987)

a. 3-9 month olds dishabituate to new basic level items
b. or habituate after several trials to presentations of basics from the same superordinate category

2. prototypes (Bomba & Siqueland, 1983)

a. 3-4 month olds continued to habituate to prototypes

C.  Types of prototype information

1.  mental average - Lasky (1974)
2.  most typical instance -  Rosch (1976)

D.  Preference vs. ability

1.  Waxman (1992)

a. children (3 & 4 yr. olds), like adults, preferred basic level labeling, but could label at other, nonbasic levels

2.  child basic level vs. adult basic level

a. e.g., in language acquisition “ball” may mean something different for children and adults (Mervis, 1987)
b. movement can be seen in 2-year olds who are given feedback

IV. Theory-based Representations (schemas)

A. Kiel (1989)

1. concepts as partial theories
2. tied to one’s associative knowledge
3. causal relations are basic
4. hierarchical relations are also informative

B. 3 core theories

1. naïve physics (nature of inanimate objects)
2. naïve biology (nature of living things)
3. naïve psychology (nature of mind)

C. age at first evidence of development

1. physics @ birth (?)
        a. e.g., Baillargeon’s work with object permanence
2. psychology @ 18 months
3. biology @ 2-3 years

I.  Themes

 A.  task analysis: e.g., math problems, working backwards

 B.  encoding: McCloskey & Kaiser (1984)

  1. misencoding of the trajectory of a falling object

 C.  mental models

  1. should accurately represent the deep structure of a problem
  2. age differences (Vosniadou & Brewer, 1992)

   a. mental models for the shape of the earth
   b. even in 5th grade 40% had non-spherical models

 D.  domain-specific vs. domain-general knowledge

  1. Chi (1978) and chess experts

 E.  developmental differences

1. Piaget suggested that young children could not create scientific concepts
2. but, even adults have difficulty (and sometimes children succeed)

A.  task analysis - weight & distance

B.  mental models (4 rules) - rule assessment method

C.  developmental differences

1. some development before age 5 (Case, 1985)

a. 4-8 mo olds: could recreate a bell ring by imitating the experimenter (involved a push down, lever-like device)
b. 12-18 mos old: could depress the lever so the other end would ring the bell
c. 2-3.5 years: could figure out the device on their own
d. 4-5 years: could use differently-weighted blocks to move the lever and ring the bell

D.  learning and encoding

1. moderate discrepancy hypothesis

a. 5 & 8 year olds given problems to solve on the next higher level (Rule II)
b. 8 year olds were more successful than 5 yr olds and often advanced to Rule III

2. specific training on encoding

a. reconstruction of the problem by 5 yr olds (Rule I users) showed difficulties with encoding
b. training on encoding of distance and weight increased their correct encoding (70% success rate)

E.  domain-general processes

1. similarities in performance on related 2-D problems

a. shadows problem (distance & size; Seigler, 1981)
b. temperature & sweetness (Strauss, 1982)
c. happiness & fairness (Marini, 1992)

A. older children learn faster & more effectively!!!

B. differences in other basic processes may help explain --> e.g., planning, causal inference, analogy use, & logical deduction