Developmental Changes in Strategies for Gathering Evidence About Biological Kinds

Abstract

How do people gather samples of evidence to learn about the world? Adults often prefer to sample evidence from diverse sources-for example, choosing to test a robin and a turkey to find out if something is true of birds in general. 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. The current study (N = 247) found that this discontinuity stems from developmental changes in standards for evaluating evidence-younger children chose to learn from samples that best approximate idealized views of what category members are supposed to be like (e.g., the fastest cheetahs), with a gradual shift across age toward samples that cover more within-category variation (e.g., cheetahs of varying speeds). These findings have implications for the relation between conceptual structure and inductive reasoning, and for the mechanisms underlying inductive reasoning more generally.

Keywords: Biological kinds; Category induction; Conceptual development; Diversity-based reasoning; Ideals.

Figure 1.

Sample procedure of trials in the Set condition. Participants completed five trials in which they worked with the character to learn about an animal kind by choosing a two-element sample set. On each trial, the character selected one exemplar on the animal scale and the participant selected a second exemplar on the scale. Across trials for each participant, the character selected one animal at each place on the scale. The participant could then select any animal on the scale they wanted except the one chosen by the character. Which exemplar the character selected for each animal kind and the order of presentation of animal kinds were counterbalanced across participants. Animal kinds included cheetahs, sharks, porcupines, giraffes, and skunks. In the Single condition, participants were asked, “Pick which one you want to look at to learn the most about [animal].” They then selected a one-element sample on their own.

Figure 2.

Participant selections on the animal scales in the Set condition, by character choice, on the animal scale. Lines represent children’s average responses across age, and the green shaded area reflects 95% confidence intervals of those averages. Large triangles reflect adults’ average responses, and error bars reflect 95% confidence intervals of those averages. Small dots reflect individual participant responses on each trial. Age is treated continuously on the x-axis, except for adults (who are graphed as a single group). Higher values are more extreme on characteristic features (e.g., faster cheetahs).

Figure 3.

The most likely strategy that each participant used in their sample selection (as predicted by the Bayesian model) in the Set condition. (A) Violin plots reflect the distribution of strategy use across age (children’s exact age is plotted as a continuous variable and adults are grouped categorically). Small dots reflect individual participants. Here, we plot only the top three strategies of all eight possible strategies, given that our model predicated that 86% of participants used one of these three strategies. (B) Frequency table of the number of participants in each age group that selected each strategy. “Other” collapses across the other five strategies listed in Section 2.1.2.