Poster #37 - The development of induction strategies in 4-7 years old children across three domains of knowledge

• In Event: 3-159 - Poster Session 12
  In Poster Session: PS 12 Section - Cognitive Process

Integrative Statement

It has been repeatedly reported that second graders’ inductive strategies exhibit signs of sensitivity to the domain of knowledge (natural kinds versus artifacts) whereas younger children do not (Gelman, 1988). However, other domains such as the food domain, have been described as borderline cases and remained under-researched so far. To fill this research gap, the present study investigated how young children’s developing inductive reasoning abilities deal with the food domain in comparison with both the natural kind and the artifact domains.
Two main research hypotheses have been tested: H1) kindergartners (4-5 years) do not exhibit inductive reasoning strategies between these domains but older ones (6-7 years) do, H2) inductive reasoning patterns are impacted by the degree of transformation of the food (i.e. untransformed, sliced, and processed/pureed). Prior to the study, caregivers filled out a validated scale (CFRS, Rioux et al., 2017) that includes two dimensions, food neophobia and pickiness, whose influence on inductive reasoning has been reported (Rioux, et al., 2018). Then, in a within-subject design (figure 1), the participants were asked to perform an inductive reasoning task. 96 children were divided into kindergartners (n=44, 4-5 years) and first graders (n=52, 6-7 years). An adult control group (n= 48) was also tested. The experimental material consisted of 54 photographs of unfamiliar foods, natural kinds and artifacts. All the exemplars were unknown to the children but a pretest confirmed that they were able to recognize the respective domains of the stimuli. Participants were told a blank property F of an inductor A and then were asked whether a target stimulus B has F or not.
To test our hypotheses, we fitted eleven multiple linear models to the data with the generalization of the property as the binary outcome measure. The best model was the three-way interaction between the inductor’s domain (food, natural kind, and artifact), the target stimulus’ domain and the age. Post-hoc Tukey comparisons, and correlation analyses revealed a clear developmental progression: younger children generalized the properties randomly and were not sensitive to the domains of the inductors and the target stimuli. On the contrary, older children exhibited systematic patterns of inference depending on these two factors. These patterns were similar to those exhibited by adults. H2 was also partially confirmed since in older children the higher degree of transformation of the target food stimuli impacted negatively the generalization of properties when the inductor was an untransformed food exemplar (figure 2). Correlation analyses revealed that inductive inferences were impacted by individual characteristics such has the intensity of food neophobia.
To conclude, the results demonstrate i) the development of domain specific inductive reasoning strategies in first graders and ii) speak in favor of the hypothesis that the food domain is a borderline case with respect to the natural kinds versus artifacts distinction. The present experiment sheds new lights on the development of the food learning mechanisms and therefore lay the ground work for designing evidence-based food education programs targeting preschoolers (Gripshover & Markman, 2013).

Authors

• Jérémie Lafraire, Institut Paul Bocuse
  Presenting Author

• Jad Hamaoui, Institut Paul Bocuse
  Non-Presenting Author

• camille rioux, Max planck institute for human development
  Non-Presenting Author

• Jean-Pierre Thibaut, Université de Bourgogne
  Non-Presenting Author