Poster #41 - Number sense outperforms general magnitude sense for both acuity and bias: Evidence from children and adults

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Cognitive Processes

Integrative Statement

Humans are thought to use the approximate number system (ANS) to make quick approximations of number important for day-to-day life (Feigenson, Dehaene, & Spelke, 2004). ANS acuity is defined as a person’s ability to correctly detect differences in the number of objects in a scene without counting them. However, differences in non-symbolic number are inevitably correlated with differences in other magnitude dimensions such as total surface area, convex hull, and density. There has long been controversy regarding the salience of number versus other stimulus dimensions throughout development. Recently, this longstanding debate has been amplified with the proposal of the sense of magnitude theory by Liebovich and colleagues (2017), which posits that number sense is not innate and is secondary to a general magnitude sense.
In the present study, we use a regression approach first described by DeWind et al. (2015) to disentangle the effects of visual magnitudes including number and total area on participant choices in a number and an area discrimination task. This approach quantifies a person’s acuity in making magnitude judgments, and the biasing effect that other magnitudes have on it. Using this novel method, we asked whether 30 five-year-old and 52 adult participants showed greater ability to discriminate number or surface area. In a nonsymbolic comparison task, half the participants were instructed to choose dot arrays based on number and half to choose based on total surface area (Figure 1).
We found that both adults and children could discriminate closer numerical than surface area ratios, reflecting better task performance in the number condition for both age groups (Figure 2). On average, adults in both conditions could discriminate closer ratios than children (Figure 2). For both children and adults, acuity and bias differed depending on task (Figure 2). Whereas adults chose based on the instructed dimension with no bias, five-year-old children chose based on the instructed dimension with significant bias. Children in the total area condition demonstrated substantial bias toward number, whereas children in the number condition only demonstrated a slight bias toward total area. On average, adults in the number condition demonstrated higher acuity than adults in the total area condition. Similarly, children in the number condition demonstrated higher acuity than children in the total area condition. These results provide evidence that number is more salient than total surface area even before the start of formal education and are inconsistent with the proposal put forth by Liebovich and colleagues (2017).

Authors

• Rachel C Tomlinson, University of Michigan
  Presenting Author

• Nicholas K DeWind, University of Pennsylvania
  Non-Presenting Author

• Elizabeth M Brannon, University of Pennsylvania
  Non-Presenting Author