Identifying distinct sources of whole number interference in children's decimal comparison: the role of numerical magnitude and inhibitory control

Decimals are challenging to learn because they introduce concepts that counter students' intuitive numerical understanding. Specifically, decimals sometimes contradict the whole number rule that longer numbers are always larger. Whole Number Bias is an overreliance on whole knowledge, leading to errors like 0.26 > 0.8. It is not well understood what cognitive processes drive overcoming this bias. First, Whole Number Bias may not solely arise from digit length interference (more digits = larger number), but also by whole referent magnitude interference (e.g., 26 > 8). Second, inhibitory control correlates with decimal performance, but it is unknown whether it suppresses one or both types of interference type. Finally, overcoming Whole Number Bias may be a radical conceptual change, or a dual process where learners gradually build decimal knowledge and suppress whole knowledge. Students in the U.S. (grades 6-8, n = 178) completed computerized decimal comparison and inhibitory control tasks. Decimal performance fell into three groups, and the two largest groups, "Whole Number Bias" and "High Performing," differed in their extent of digit length interference, rational magnitude processing, and inhibitory control, aligning with conceptual change theory. However, in High Performing only, inhibitory control correlated with less digit length interference, aligning with dual process theory. Unexpectedly, both groups were comparably susceptible to whole referent interference, and inhibitory control did not correlate with such susceptibility. Our findings suggest learning decimals involves first a discontinuous shift in conceptual understanding, then gradual improvement where inhibitory control supports overcoming digit length but not whole referent interference.