Overcoming appearances: Teleological essentialism in children's categorization

This repository contains the experiments, data, analyses, and figures for the paper "Overcoming appearances: Teleological essentialism in children's categorization".

Contents:

• Introduction
• Repository structure

Introduction

Overcoming appearances in categorizing is an important intellectual achievement. But children often rely heavily on appearances, as classic shape similarity (e.g., Landau et al., 1988) and perceptual transformation tasks (Keil, 1989) demonstrate. We propose that children can use purposes to overcome appearances when categorizing not only artifacts, but biological kinds as well. Across three experiments, with 570 US-based children, we find support for this proposal. Children successfully categorized dissimilar-looking objects together based on shared purposes, both at basic (Experiment 1) and superordinate levels (Experiment 2), and these effects were comparable for artifacts and biological kinds. In Experiment 3, which focused on objects undergoing radical transformations, children also used an object’s purpose to judge that it maintained its identity in the face of radical perceptual transformations, and this improvement was at least as strong for biological kinds as for artifacts.. We discuss our results in the context of essentialism. Some have proposed that scientific essences – placeholders for what gives something its identity that science fills out – might help children overcome appearances, but this requires a protracted development since scientific elaboration is challenging and rare (Gelman, 2003). As one alternative, we suggest that children can readily treat purposes as identity-relevant, and that this may reflect an early teleological construal of what makes something the kind of thing it is.

Repository structure

├── code
│   ├── R
│   └── experiments
│       ├── experiment1
│       ├── ...
├── data
│   ├── experiment1
│   ├── experiment2
│   └── experiment3
├── docs
└── figures
    ├── experiment1
    ├── experiment2
    ├── experiment3

• code/ contains all the code for the experiments, analyzing data and generating figures.
• experiments contains materials and code for each experiment that was run. Pre-registrations for all experiments are linked below.
  • experiment1 (pre-registration) was run asynchronously in Lookit.
    • experiment2 (pre-registration) was run synchronously over Zoom.
  • experiment3 includes a two-part norming study and the experiement
    • two part norming study: - part1 (pre-registration) and part2 (pre-registration) were run asynchronously in Lookit.
      • experiment:
    • experiment (pre-registration) was run asynchronously in Lookit.
• R contains the analysis scripts that were used to analyze data and generate figures.
• data/ contains anonymized data from all experiments:
  • experiment1 contains exp1.csv which includes trial and demographic data.
  • experiment2 contains exp2.csv which includes the trial and demographic data.
  • experiment3 contains exp3_norm_part1.csv, exp3_norm_part2.csv, adn exp3.csv which includes the trial demographic data for each respective experiement.
• docs/ contains R code for viewing
• figures/ contains all the figures from the paper (generated using the script in code/R/).