PROJECT SUMMARY / ABSTRACT Basic mathematical skills, sometimes referred to as functional numeracy, are essential for successful everyday living and are associated with important social determinants of health (e.g., employment, job promotion, income, financial literacy, health literacy)1–3. African American children, many of whom are speakers of a non- mainstream dialect of American English4–6, are less likely to attain basic mathematical competence than their White peers7–9 and are more likely to struggle to attain functional numeracy. One promising, yet largely unexplored, way of understanding disparities in the attainment of functional numeracy, specifically, is to examine arithmetic strategy usage in young children. Children's selections of strategies and their patterns of errors during the execution of strategies are important indicators of the development of functional numeracy10– 15. Both child abilities (e.g., working memory) and item features (e.g., problem-size) appear to predict children's arithmetic strategy usage10–15,13,14,16–19,17,20. However, research has not examined both child abilities and item features in comprehensive measurement models, nor has it examined the possibility that African American children may encounter unique problem demands as a result of cultural-linguistic factors. Recent advances in measurement theory and available statistical software present researchers with the opportunity to synthesize and expand current understandings of children's strategy selections using advanced latent variable modeling techniques21–27. These techniques are particularly suited for examining test validity, an issue which is particularly important for understanding the arithmetic strategy usage patterns of children who may encounter measurement bias as a result of cultural factors. The overall goal of the proposed project is to examine measurement models of children's arithmetic strategy usage with an extant dataset of N=872 children, and to extend these models toward understanding the unique mathematics learning challenges of African American children during a sensitive period for the identification of mathematics learning difficulty (i.e., the transition period between elementary and middle grades, during late childhood and early adolescence)28,29 The proposed postdoctoral NRSA has the following specific aims and associated training objectives: Specific Aim 1: To use extant cognitive and behavioral datasets to test explanatory models of relations between children’s arithmetic strategy usage and mathematical competence in year 1. Associated Training: Achievement of this aim will be facilitated by continued training in the theories and applications of both arithmetic cognition and psychometrics. Specific Aim 2: To conduct pilot testing with N=100, African American children who use African American English dialect (AAE) in years 1 - 3, building on results from initial measurement models examined in Aim 1. Associated Training: Achievement of this aim will be facilitated by ongoing mentorship in psychometric theory and measurement of arithmetic cognition, as well as mentorship in community-based research with African- American populations.

PROJECT NARRATIVE Young children's arithmetic strategy usage is a promising indicator of mathematical learning difficulty; however, current approaches to measuring and modeling arithmetic strategy usage leave unanswered questions about (1) how children's abilities contribute to strategy usage, and (2) why their strategies may vary across different types of items. These theoretical issues have important implications for African American children, many of whom may struggle with arithmetic word problems as a result of linguistic mismatch with their cultural dialect. The proposed study will help to inform the interpretation of arithmetic assessments for a variety of children, in effect, laying the groundwork for a more nuanced identification of children's mathematics learning profiles.