When proportional information is pit against whole number numerical information, children often attend to the whole number information at the expense of proportional information (e.g., indicating 4/9 is greater than 3/5 because 4 > 3). In the current study, we presented younger (3- to 4-year-olds) and older (5- to 6-year-olds) children a task in which the proportional information was presented either continuously (units cannot be counted) or discretely (countable units; numerical information available). In the discrete conditions, older children showed numerical interference-responding based on the number of pieces instead of the proportion of pieces. However, older children easily overcame this poor strategy selection on discrete trials if they first had some experience with continuous, proportional strategies, suggesting this prevalent reliance on numerical information may be malleable. Younger children, on the other hand, showed difficulty with the proportion task, but showed evidence of proportional reasoning in a simplified estimation-style task, suggesting that younger children may still be developing their proportional and numerical skills in task-dependent ways. Lastly, across both age groups, performance on the proportional reasoning task in continuous contexts, but not discrete contexts, was related to more general analogical reasoning skills. Findings suggest that children's proportional reasoning abilities are actively developing between the ages of 3 and 6 and may depend on domain general reasoning skills. We discuss the implications for this work for both cognitive development and education. (PsycINFO Database Record

No abstract

When placing numbers along a number line with endpoints 0 and 1000, children generally space numbers logarithmically until around the age of 7, when they shift to a predominantly linear pattern of responding. This developmental shift of responding on the number placement task has been argued to be indicative of a shift in the format of the underlying representation of number (Siegler & Opfer, ). In the current study, we provide evidence from both child and adult participants to suggest that performance on the number placement task may not reflect the structure of the mental number line, but instead is a function of the fluency (i.e. ease) with which the individual can work with the values in the sequence. In Experiment 1, adult participants respond logarithmically when placing numbers on a line with less familiar anchors (1639 to 2897), despite linear responding on control tasks with standard anchors involving a similar range (0 to 1287) and a similar numerical magnitude (2000 to 3000). In Experiment 2, we show a similar developmental shift in childhood from logarithmic to linear responding for a non-numerical sequence with no inherent magnitude (the alphabet). In conclusion, we argue that the developmental trend towards linear behavior on the number line task is a product of successful strategy use and mental fluency with the values of the sequence, resulting from familiarity with endpoints and increased knowledge about general ordering principles of the sequence.A video abstract of this article can be viewed at:http://www.youtube.com/watch?v=zg5Q2LIFk3M.

Although fractions, decimals, and whole numbers can be used to represent the same rational-number values, it is unclear whether adults conceive of these rational-number magnitudes as lying along the same ordered mental continuum. In the current study, we investigated whether adults' processing of rational-number magnitudes in fraction, decimal, and whole-number notation show systematic ratio-dependent responding characteristic of an integrated mental continuum. Both reaction time (RT) and eye-tracking data from a number-magnitude comparison task revealed ratio-dependent performance when adults compared the relative magnitudes of rational numbers, both within the same notation (e.g., fractions vs. fractions) and across different notations (e.g., fractions vs. decimals), pointing to an integrated mental continuum for rational numbers across notation types. In addition, eye-tracking analyses provided evidence of an implicit whole-number bias when we compared values in fraction notation, and individual differences in this whole-number bias were related to the individual's performance on a fraction arithmetic task. Implications of our results for both cognitive development research and math education are discussed. (PsycINFO Database Record

Recent research suggests that fraction understanding is predictive of algebra ability; however, the relative contributions of various aspects of rational number knowledge are unclear. Furthermore, whether this relationship is notation-dependent or rather relies upon a general understanding of rational numbers (independent of notation) is an open question. In this study, college students completed a rational number magnitude task, procedural arithmetic tasks in fraction and decimal notation, and an algebra assessment. Using these tasks, we measured three different aspects of rational number ability in both fraction and decimal notation: (1) acuity of underlying magnitude representations, (2) fluency with which symbols are mapped to the underlying magnitudes, and (3) fluency with arithmetic procedures. Analyses reveal that when looking at the measures of magnitude understanding, the relationship between adults' rational number magnitude performance and algebra ability is dependent upon notation. However, once performance on arithmetic measures is included in the relationship, individual measures of magnitude understanding are no longer unique predictors of algebra performance. Furthermore, when including all measures simultaneously, results revealed that arithmetic fluency in both fraction and decimal notation each uniquely predicted algebra ability. Findings are the first to demonstrate a relationship between rational number understanding and algebra ability in adults while providing a clearer picture of the nature of this relationship.