Constructing causal stories

This repository contains the experiments, data, analyses, and figures for the paper "Constructing causal stories: How mental models shape memory, prediction, and generalization".

• Constructing causal stories
  • Introduction
  • Repository structure
    • code
      • blender
      • python
      • R
    • data
    • docs
    • figures
  • Pre-registrations
    • Experiment 1
    • Experiment 2
    • Experiment 3
  • Experiment demos
    • Experiment 1
    • Experiment 2
    • Experiment 3
  • CRediT author statement

Introduction

How do people learn to predict what happens next? On one account, people do so by building mental models that mirror aspects of the causal structure of the world. Accordingly, people tell a story of how the data was generated, focusing on goal-relevant information. On another account, people make predictions by learning simple mappings from relevant features of the situation to the outcome. Here, we provide evidence for the causal account. Across three experiments and two paradigms, we find that people misremember what happened, predict incorrectly what will happen, and generalize to novel situations in a way that's consistent with the causal account and inconsistent with a feature-based alternative. People spontaneously construct causal models that compress experience to privilege causally relevant information. These models organize how we remember the past, predict the future, and generalize to novel situations.

Repository structure

.
├── code
│   ├── blender
│   ├── python
│   └── R
├── data
│   ├── experiment1
│   ├── experiment2
│   └── experiment3
├── docs
│   ├── analysis
│   ├── experiment1
│   ├── experiment2
│   └── experiment3
└── figures
    ├── diagrams
    ├── plots
    └── stimuli

code

blender

• Blender files used for rendering 3D stimulations.
• See code/blender/README.md for details.

python

• Python files used for running physical simulations + implementation of the causal abstraction model.
• See code/python/README.md for details.

R

• R files used for data wrangling, visualization, and for fitting the features model.
• causal_stories.Rmd depends on modeling results produced with python (specifically, it reads in model_vs_human_comparison.csv).

data

Raw data files from each experiment.

docs

• analysis/ has the rendered RMarkdown document which can also access here.
• The experiment/ code folders contain the code for each experiment. See the readme files within the experiment folders for additional information. You can try out demos of the experiments by clicking on the links in Experiment demos.

figures

All the figures in the paper and experiment stimuli.

Pre-registrations

Experiment 1

• feedback condition
• no feedback condition
• short condition
• conjunctive condition

Experiment 2

• long condition
• short condition
• conjunctive condition

Experiment 3

• both conditions

Experiment demos

Links to demos of the different experiments.

Experiment 1

• feedback condition
• no feedback condition
• short condition
• conjunctive condition

Experiment 2

• long condition
• short condition
• conjunctive condition

Experiment 3

• forward facing
• backward facing

CRediT author statement

For details about the different terms see here.

Term	Steven	Chuqi	Paul	Tobi
Conceptualization	x		x	x
Methodology	x	x		x
Software	x	x		x
Validation	x	x		x
Formal analysis	x			x
Investigation	x
Data Curation	x	x
Writing - Original Draft				x
Writing - Review & Editing	x	x	x	x
Visualization				x
Supervision			x	x
Project administration	x			x
Funding acquisition				x