Does math education modify the approximate number system? A comparison of schooled and unschooled adults

Does math education contribute to refine the phylogenetically inherited capacity to approximately process large numbers? The question was examined in Western adults with different levels of math education. Unschooled adults who never received math education were compared to unschooled-instructed adults who did not attend regular school but received math education in adulthood, and to schooled adults who attended regular school in childhood. In the number-comparison task (Exp. 1), the unschooled group was slower and made more errors than the other groups both when numerical symbols and nonsymbolic dot collections were presented. In the forced-choice mapping task (Exp. 2), the unschooled group experienced more difficulty than the others in linking large nonsymbolic and symbolic quantities, as well as in matching purely nonsymbolic quantities. These results suggest that Western adults who did not receive math education have less precise approximate number skills than adults who acquired exact number competences through math education.