Poster #13 - Recognizing small numbers; the developmental trajectory of cardinal knowledge in early childhood

• In Event: Poster Session 13

Abstract

Being able to understand numbers and what they represent is an important part of school readiness (Del Rio et al., 2017), academic achievement (Dunlan et al., 2007), and success in several areas of adult life (Smith et al., 2015). Thus, the importance of numerical understanding and tools to foster it in childhood cannot be overstated. At the same time, it is still unclear and debated exactly how children come to understand number words.
It is a common practice to use the give-n task to group children into discrete knower levels; either a pre-knower, subset- knower or cardinal-principal knower. Recent research suggest that subset- knowers have a partial understanding of small numbers (i.e. 1-4), and might use this approximate knowledge to develop an exact understanding of numbers (O ’Rear et al, 2020; Wagner et al, 2019). However, there is some conflicting evidence of what partial knowledge subset-knowers possess. Using a point-to-X task, Silver et al (2021) found that one- knowers could readily identify targets ranging 1- 4. Contrary, Pinhas et al (2014) demonstrated that one- knowers did not react when number words were used incorrectly in a mismatch task.
As to address this conflicting evidence, we developed a new eye-tracking task. Participants were presented with two sets differing in cardinality (1–4 items) and asked to find a specific cardinality. This allows us to present more trials (48 in total), control for that the target number is presented equal times, and showing all possible combinations of numbers equal times, as well as making sure the conditions are more similar for each participant. Finally, it allows for measuring first gaze, capturing a more instantaneous response of cardinal recognition, which we argue should capture children’s (partial) knowledge more readily.
Our sample consisted of 133 participants (middleclass background with a majority born in Sweden), aged 2.6 – 4 years (Mage = 3. 4 years, SDage = .6, min = 2. 6, max = 4. 7; 55% female). Using t-tests, we demonstrate that, on a group level, subset knowers could identify numbers 1- 4 reliably (p < .01), further demonstrating a partial understanding of number words in early childhood. However, using a mixed model, we also found that knower level affected the performance (F(5,138) = 4.70, p <.001). Exploratory t-tests analysis suggest that both pre- and one-knowers performed at chance (p > .05) on every target number, whereas two- up to cardinal principal knowers performed above chance (p > .001). Thus, in this study, it seems that one-knowers could not even identify “one” reliably, in line with Pinhas et al (2014). It might be that these children see “one” as a linguistic marker, interchangeable with “a/an/the”, instead of a cardinal marker.
Although future research is needed to make definitive conclusions, we tentatively suggest an important conceptual leap between one-knowers and two-knowers; that when children have learned what “two” refers to, they can also comprehend and identify the terms 1-4 reliably. If true, it would suggest that understanding of number words actually begin at the two-knower level.

Author

• Anton Gerbrand
  Presenting Author