Understanding and reducing left digit bias in numerical magnitude judgments

The National Science Foundation Division of Research on Learning in Formal and Informal Settings awarded Wesleyan University $736,381 on August 1, 2026, to examine left digit bias in numerical magnitude judgments and develop interventions to reduce it under the STEM Education program (CFDA 47.076).

The project will conduct ten studies investigating how people disproportionately weight the leftmost digit when estimating the size of numbers—a bias that emerges by age nine and persists into adulthood and that is revealed through number line estimation tasks used in K–12 mathematics classrooms. Studies 1 through 3 will use correlational approaches with college students and 9- to 14-year-old children to determine whether left digit bias in number line estimation predicts other mathematical skills and whether it predicts similar bias in everyday numerical judgment and decision-making contexts. Studies 4 through 6 will investigate whether left digit bias is malleable or fixed once it emerges. The project will also develop and test a training intervention aimed at reducing the bias, with potential applications to mathematics learning.

Performance occurs in Middletown, Connecticut. The period of performance runs through July 31, 2029.

Abstract

Numbers are central to everyday life, from reading a price tag to making a medical decision. When judging the size of numbers, people often make errors that follow a predictable pattern. For example, research has shown that the leftmost digit of a number exerts a disproportionate influence on how people estimate its magnitude, a phenomenon known as left digit bias. The bias is most clearly revealed through number line estimation tasks, widely used both to study how children develop number sense and as instructional tools in k-12 mathematics classrooms. Understanding left digit bias is therefore important both for mathematics education and for improving everyday judgment and decision making across many domains. This project will examine how the bias relates to other number skills and math competencies in children and adults, whether IT is fixed once IT emerges or can be changed, and what mental processes underlie IT. The project will also develop and test a training intervention aimed at reducing left digit bias, with potential benefits for mathematics learning. Symbolic number line estimation tasks, in which target numbers are placed on lines bounded only at their endpoints, reveal a robust bias in estimation called a left digit bias. The bias, which reflects a disproportionate influence of leftmost digits of target numerals on placements, emerges by 9 years of age and extends through adulthood. Ten studies will be conducted to understand the left digit bias and to test a training intervention developed to reduce bias. Studies 1 through 3 will use a correlational approach with college students and 9-14-year-old children to address whether left digit bias in number line estimation predicts other mathematical skills and whether IT predicts a similar bias in everyday number-related judgment and decision-making contexts. Studies 4 through 6 will probe the malleability of the left digit bias, using a cross-linguistic adult internet sample to assess whether bias is smaller for languages with an inversion property (digits spoken from right to left); and separately, assessing the extent to which children (like adults) flexibly show bias for the most informative leftward digit even if IT is not leftmost. Studies 7 through 9 will use mathematical modeling methods with existing and new data from adults and children to test competing accounts of the cognitive source of the bias: by a weighting account, bias arises from the relative underweighting of rightward digits, while by a digit bias account IT arises from the extension of a bias in whole-number magnitude estimation to each individual digit. In study 10, an intervention will be developed and tested in college students, aimed at reducing left digit bias by implicitly directing attention to rightward digits, with findings that will inform both theoretical accounts of the bias and practical approaches to improving numerical estimation. This project was funded by the stem k-12 program, which supports fundamental, applied, and translational research on stem learning in both formal k-12 education and informal learning environments.? This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.- subawards are not planned for this award.