[DSRP Evidence](https://dsrpevidence.org/)

# A Theory of Causal Learning in Children: Causal Maps and Bayes Nets

## Details

**Authors** Gopnik et al.

**Year** 2004

**Publisher** Psychological Review

**Discipline** Human Development

[Read it at the publisher](https://doi.org/10.1037/0033-295X.111.1.3) 
10.1037/0033-295X.111.1.3

## In authors' words

### Abstract

The authors outline a cognitive and computational account of causal learning in children. They propose that children use specialized cognitive systems that allow them to recover an accurate "causal map" of the world: an abstract, coherent, learned representation of the causal relations among events. This kind of knowledge can be perspicuously understood in terms of the formalism of directed graphical causal models, or Bayes nets. Children's causal learning and inference may involve computations similar to those for learning causal Bayes nets and for predicting with them. Experimental results suggest that 2- to 4-year-old children construct new causal maps and that their learning is consistent with the Bayes net formalism.

### What they set out to do (purpose)

To explore the causal structure of the world and how children use that structure to learn content that has typically a steep learning curve.

### What they found (results)

They hypothesized that,“children use specialized cognitive systems that allow them to recover an accurate “causal map” of the world: an abstract, coherent, learned representation of the causal relations among events.” Their experiments indicated that children ages 2 to 4 years old were able to construct such causal maps, and their learning process was similar to the “Bayes net formalism” in which they saw far more than simple one way relationships among things.

## Commentary

### In short

Children are fundamentally able to build complex relationships and utilize the Rar structure.

### In more detail

Gopnik et. al. 2004 explored the causal structure of the world and how children use that structure to learn content that has typically a steep learning curve. In their research, they hypothesized that,“children use specialized cognitive systems that allow them to recover an accurate “causal map” of the world: an abstract, coherent, learned representation of the causal relations among events.” They found that a possible method for causal learning and inference in children is computations that resemble the learning and prediction process for Bayes nets, which simply put, are representations of multiple variables and their dependencies (or a web of causality). Their experiments indicated that children ages 2 to 4 years old were able to construct such causal maps, and their learning process was similar to the “Bayes net formalism” in which they saw far more than simple one way relationships among things.

**Patterns it shows** R

**How to cite this** Gopnik et al. (2004). A Theory of Causal Learning in Children: Causal Maps and Bayes Nets. Psychological Review.