Lecture 12 - Reasoning, Exploration, Motivation

What was Piaget’s hypothesis on children’s abstract reasoning?

Children develop in stages. Abstract reasoning at age 11-12.

Define analogical reasoning.

Comparing things that on the surface seem different in order to see deeper-level similarities between them

ex.

Superficial: color, shape

Deeper: sameness, two of the same

What age do children do analogical reasoning?

By age 6 children can solve these problems.

• 3-4 yo fail at tasks but succeed if it involves casual reasoning and social cognition.

• evidence of analogical reasoning in younger kids if question framed causally

What do we mean by “Child as scientist”?

Scientist and Children:

1. Theories

2. Isolate candidate causes

3. Infer causal relationships from patterns in stats

4. Constrain generalizations depending on evidence sampled

5. How much evidence necessary to draw conclusion

EVIDENCE OF THEORIES AND CAUSES: Do children distinguish between competing causes in infancy?

Specific reactions to specific violation of expectation (i.e. testing material properties)

EVIDENCE OF THEORIES AND CAUSES: Do children explore when their theory changes? 

Step 1. measure whether child is geometric or mass theorist

Step 2. random assignment to geometric center or center of mass condition (confirms or disconfirms theory of balance)

Step 3. free choice between original toy or novel toy

Question: Do children explore differently depending on whether their theory is supported or violated?

Belief consistent = choose new toy

Belief inconsistent = explore original toy

EVIDENCE OF THEORIES AND CAUSES: Do children explore when evidence is confounded?

Confounded = unclear which lever cause effect (ambiguity)

Unconfounded = clear which level cause effect

• presented play choice of same toy or new toy

Ambiguity in causal system motivated children to explore

EVIDENCE OF INFER CAUSAL RELATIONS: Do children infer underlying causes from seeing statistical patterns?

• within-agents condition: toy works half the time

  • “broken toy” = change toy

• between-agents condition: toy works with specific people 

  • “person-specific” = change agent

• 18mo either change object (new toy) or agent (give to parent)

18mo infer causes of their own failures

EVIDENCE FOR SAMPLE GENERALIZATION: Do children do intuitive statistics?

Step 1. experimenter drew 5 marbles from box

Step 2. population revealed (either expected or unexpected)

Step 3. measure looking times in 8mo

• tested if sample from source not of interest (i.e. pocket) = no generalization on population of box

• same effect if shown population and predict likely sample

Children relationship between sample and its population. Expect sample randomly drawn to represent population (vice versa).

EVIDENCE FOR SAMPLE GENERALIZATION: Are children sensitive to the sampling process?

Step 1. adult draws 3 blue toy and squeaks them

Step 2. baby given a yellow toy 

Step 3. does baby squeak toy? 

• Weak sampling = data randomly generated

  • ex. mostly blue box, consistent with sampling

  • more squeezes

• Strong sampling = data generated intentionally 

  • ex. mostly yellow box, cherry-picked sampling

  • less squeezes

Infants generalize property (squeak) to new toy from same population if its plausible evidence was RANDOMLY generated compared to cherry picked.

Define Intuitive Power Analysis.

To distinguish between two similar hypotheses (e.g. 50/50 vs. 51/49), you need a bigger sample size

EVIDENCE FOR SAMPLE SIZE: Are children able to present how much information they might need to solve a problem?

Children request more data for more difficult discriminations

How to teach exploration?

Direct pedagogy restricts exploration.

ex. “Let me show you how this works” vs “Huh, what’s this? Let’s see how it works”

How to motivate?

Paying children to do something they already love undermines their intrinsic motivation

• praising effort rather than person

What is children’s perspective on “who is really, really smart?” (gender stereotypes)?

“who is really, really smart?”

children age 6+ have a male = brilliant bias

What is children’s gender stereotypes for interest in STEM?

Children endorse stereotype that girls are less interested than boys in stem

• Girls with stronger stereotype = less interest they expressed in STEM

• Boys with stronger stereotype = more interest they expressed in STEM

Infants and young children engage in practices parallel to those in formal science but formal education….

• focuses on concrete facts

• direct teaching

• use explicit rewards (grades)

• gender stereotypes