Inhibitory Control Supports Young Children’s Learning of Counterintuitive Scientific Ideas

• In Event: Domain-General Mechanisms of Science Learning and Reasoning

Abstract

Scientific explanations are often deeply counterintuitive. Children’s untutored understandings of natural phenomena therefore often conflict with scientifically accurate ideas. How do children learn scientifically accurate explanations in the face of intuitive misconceptions? One possibility is that children revise and replace incorrect ideas with new, correct ideas (e.g., Carey, 1985; Piaget, 1929). Another possibility is that as children learn new explanations, their intuitions persist and co-exist alongside those new ideas (e.g., Kelemen, 2004; Shtulman, 2018). These differing accounts of conceptual change lead to distinct predictions about the role of executive functions (EF) in learning scientific concepts. If—as suggested by the conceptual co-existence account—intuitive ideas persist and compete with scientifically-accurate explanations, individual differences in EF skills, particularly inhibition, should not only predict children’s ability to initially construct scientifically accurate explanations but also their ability to subsequently express these accurate explanations after they have been learned. This study tested this prediction by examining the role of EF in the construction and expression of natural selection, a counterintuitive scientific concept that is subject to teleological and essentialist misconceptions.

Second and third graders (N=44; Mage = 8 years, SD = 7 months) participated in a storybook-based intervention designed to scaffold a causal-explanatory understanding of natural selection (e.g., Kelemen, 2019). Over several days, children completed a pretest assessment, participated in a multi-storybook intervention that explained small- and large-scale evolutionary change, and completed multiple posttest assessments to measure their ability to construct and generalize an accurate understanding of natural selection. Children’s ability to retain this generalized understanding was tested again after a three-month delay. Children also completed assessments of working memory (WM: digit span) and inhibitory control (flanker). To isolate the effects of inhibitory control, we controlled for WM and age in all analyses.

We first examined whether EF skill predicted whether children could construct an accurate understanding of natural selection after instruction (i.e., progress from not understanding natural selection at pretest to understanding it at posttest). Converging with prior research (Zaitchik et al., 2013), logistic regression revealed that participants’ inhibitory abilities predicted accurate learning of natural selection over and above effects of WM and age, p = .032. Next, we examined whether—among children who demonstrated accurate learning on immediate generalization post-tests (N=28)—inhibitory control predicted children’s ability to inhibit competing intuitive explanations on the 3-month delayed generalization posttest. Logistic regression indicated inhibition was a significant predictor over and above the effects of WM and age, p = .041.

These results are consistent with the conceptual co-existence account of conceptual change. Children’s inhibitory control capacities uniquely predicted their ability to construct a scientifically-accurate understanding of natural selection and to enduringly express it. This suggests that individual differences in specific EF skills are critical for learning counterintuitive scientific concepts because children must suppress their intuitions about the origins of organisms’ specialized traits in order to initially learn the correct mechanism and to later express that correct mechanism without the insertion of persistent, co-existent, intuitive misconceptions. Implications for conceptual change theories and education will be discussed.

Authors

• Samuel Ronfard, University of Toronto
  Non-Presenting Author

• Sarah A Brown, Boston University
  Non-Presenting Author

• Erin Doncaster, Boston University
  Non-Presenting Author

• Emma Pitt, Northeastern University
  Non-Presenting Author

• Deb Kelemen, Boston University
  Presenting Author