Developmental Changes in Strategies for Learning About Biological Kinds

• In Event: 3-134 - The Essence of Biological Categories: Factors Influencing Children’s Sensitivity to Variability and Change

Integrative Statement

Human cognition is remarkably powerful—from limited samples of evidence (e.g., two birds observed on a particular day), one can make a range of inferences about an entire category. Not all samples are equivalently informative, however. How do people determine which samples license generalizations? Adults often prefer evidence from diverse sources—for example, choosing to test a robin and a turkey to find out if something is true of all birds (Gutheil & Gelman, 1997; Lopez, Gelman, Gutheil, & Smith, 1992). Children below age 9, however, often do not consider sample diversity, instead treating non-diverse samples (e.g., two robins) and diverse samples as equivalently informative (Rhodes, Gelman, & Brickman, 2010; Rhodes & Liebenson, 2015). The present study tests a new hypothesis regarding why children's reasoning strategies change across development—that the structure of children’s concepts of the biological world leads them to apply a different standard for evaluating samples of evidence.

Recent research has found that younger children’s biological concepts are centered on ideal forms—a stylized representation of what they think members of natural categories should be like (Foster-Hanson & Rhodes, 2018). In this work, younger children viewed category members that displayed relatively extreme levels of characteristic features (e.g., the very fastest cheetah, the most stinky skunk) as the most representative and informative about their kinds, with a gradual shift across childhood towards viewing more average category members in this way. With this in mind, the present study tested the possibility that younger children’s emphasis on what category members should be like could lead them to view samples that most closely approximate these ideal forms as most informative for learning about the kind, with a shift towards preferring evidence that covers more of the variation that exists across diverse members of a category with age.

We introduced children (ages 5-10 years) and adults (N = 226) to familiar animals varying along one property dimension (e.g., cheetahs running at varying speeds). On each animal trial, the participant worked with the experimenter to choose a sample of two exemplars from which to learn about the kind. The experimenter chose the first element in each sample, and the participant chose the second. In line with the hypothesis that standards for evaluating evidence might shift across development, younger children chose to learn from exemplars that best approximate idealized views of what category members are supposed to be like (e.g., the two fastest cheetahs); sampling strategies shifted gradually across development towards covering more within-category variation (cheetahs of varying speeds). Implications for the relation between conceptual representations and inductive reasoning strategies will be discussed.

Authors

• Emily Foster-Hanson, New York University
  Presenting Author

• Kelsey Moty, New York University
  Non-Presenting Author

• John Daryl Ocampo, New York University
  Non-Presenting Author

• Amanda cardarelli, New York University
  Non-Presenting Author

• Marjorie Rhodes, New York University
  Non-Presenting Author