Does highlighting set size facilitate number learning?

• In Event: New Developments in the Science of Numerical Thinking

Abstract

Young children slowly and effortfully learn the meaning of the first few number words before they acquire an understanding of counting and cardinality (Wynn, 1992). Learning the meaning of number words may be particularly challenging because they are abstract. For example, although three yellow stars and three green flowers look nothing alike, they can both be labelled ‘three’. The abstract meaning of the numerosity “three” can be inferred from the shared property of set size between these examples. However, young children pay more attention to properties like shape and color rather than numerosity when looking for commonalities between sets (Chan & Mazzocco 2017; Mazzocco et al., 2020). Could we use sets with contrasting visual properties and numerosity to highlight that number words refer to numerosity and facilitate young children’s learning of the first number words?
We created a brief exposure for subset knowers to train their knowledge of the next number word. In one training condition, we showed examples of sets that had contrasting visual properties designed to highlight numerosity as the referent of the number word (e.g. yellow stars and green flowers). In a control training, we showed examples of sets with identical visual properties (e.g. only yellow stars). We trained 2- and 3- knowers (N = 65) on the next number word (i.e. “three” or “four”) and assessed their learning with a 2-Alternative-Forced-Choice task and Give-a-Number task. We calculated a one-sided independent samples t-test to test whether children who received the training with visual contrast would be more accurate than the children who received the control training at selecting the trained numerosity between the 2 alternative choices. This analysis was preregistered.
Children who received the visual contrast training (N = 31, M = 59.3%, SD = 21.2%) performed significantly better than the children in the control condition (N = 34, M = 49.3%, SD = 20.4%), t(63) = 1.941, p = .028, d = .482.
Children who received the visual contrast training were also better at creating sets of the trained numerosity in the Give-a-number task. Over half (N = 15, 56%) of the children who received the visual contrast training could give the trained numerosity correctly on at least two of three attempts compared to only about a third (N = 8, 32%) of the children who received the control training.
We conclude that young children may benefit from examples with contrasting properties that highlight numerosity when learning the first number words.

Authors

• Rebecca Merkley, Carleton University
  Non-Presenting Author

• Theresa Elise Wege, Loughborough University
  Presenting Author

• Pierina Cheung, Nanyang Technological University
  Non-Presenting Author

• Daniel Ansari
  Non-Presenting Author