Measuring Emerging Number Knowledge in Toddlers

• In Event: Poster Session 08
  In Poster Session: PS 08 Topic - Cognitive Processes

Abstract

Recent evidence suggests infants may recognize counting as numerically relevant long before they are able to count or understand the cardinal meaning of number words themselves (Wang & Feigenson, 2019). Children’s cardinal number knowledge is typically measured via the Give-N task that asks children to produce sets of objects in different quantities (Wynn, 1990). However, Give-N may not capture children’s approximate understanding of number words. Here, we asked whether toddlers are able to correctly map number words to the correct quantities in a forced choice task. Thirty-two 2-year-old toddlers (Mage=32 months; 14 females) completed a traditional Give-N task to measure developing cardinality knowledge and a novel Point-to-X task (based on Wynn, 1992) to examine children’s sensitivity to number words prior to understanding their exact meanings. To familiarize children with the Point-to-X task, children were first shown two word-control trials with two different common objects and were prompted to point to one image (e.g., “Where is the ball?”), which all children completed correctly. Subsequently, in twelve number-word trials, each image showed two sets of identical stimuli differing only in number (e.g., four ducks and five ducks), and children were prompted to point to one of the images (e.g., “Which has four?”). Number-word trials varied along two distinct dimensions: the numerical distance between the two sets (for “one-away” trials, the numbers differed by one; for “two-away” trials, the numbers differed by two; for “far-away” trials, the numbers differed by more than four) and size of the target number (for eight trials the correct number was small (1-4), and for four trials the correct number was large (>4); see Figure 1 for examples of each).
Across all number-word trials, children pointed to the correct image more often than predicted by chance (t(31)=3.41, p=.002; see Table 1), indicating that toddlers had some understanding of the prompted number word that allowed them to rule out incorrect responses, despite limited understanding of exact cardinal values. Furthermore, children’s performance was not associated with their age or sex, and counter to hypotheses, performance did not differ for trials of different numerical distances. However, performance was significantly better and children pointed to the correct quantity more often for trials where the target number was small compared to trials where the target number was large (t(31)=2.10, p=.044). Finally, children with better understanding of exact number words as indicated on the Give-N task answered more trials correctly in Point-to-X (=.49, p=.003). As researchers move to developing assessments that can be administered remotely, Point-to-X promises to be an easy-to-administer alternative to Give-N for measuring children’s emerging number knowledge. In fact, pilot testing with 14 2-year-olds (Mage=34 months; 9 females) using an online version of the task indicates no significant difference in children’s performance when tested in person versus remotely, supporting its efficacy in remote testing. Thus, the novel Point-to-X measure seems to be a sensitive measure for capturing children’s developing number word knowledge both in person and remotely.

Authors

• Alex M. Silver, University of Pittsburgh
  Presenting Author

• Emily Braham, University of Pittsburgh
  Non-Presenting Author

• Leanne Elliott, University of Pittsburgh
  Non-Presenting Author

• Catherine S Tamis-Lemonda, New York University
  Non-Presenting Author

• Natasha J Cabrera, University of Maryland - College Park
  Non-Presenting Author

• Heather J. Bachman, University of Pittsburgh
  Non-Presenting Author

• Elizabeth Votruba-Drzal, University of Pittsburgh
  Non-Presenting Author

• Melissa E Libertus, University of Pittsburgh
  Non-Presenting Author