The Language of Numbers: Reading Comprehension and Applied Math Problem-Solving

Column Descriptions and Encodings:<b>Student_id</b>: Unique integer identifier for each participant (e.g., 1, 3, 4, ...). This is an arbitrary ID assigned for data organization and does not carry additional meaning.<b>gender_f1m2</b>: Gender of the participant, encoded as 1 = female, 2 = male. This demographic variable was collected to describe the sample but was not a primary focus in analyses.<b>Age(Y.M)</b>: Age of the participant in years and months (e.g., 9.3 = 9 years and 3 months; 16.5 = 16 years and 5 months). Ages reflect the developmental stages: fourth graders (Group 1) range from approximately 9-10 years; eleventh graders (Group 2) range from approximately 16-17 years.<b>Group</b>: Age/grade group of the participant, encoded as 1 = fourth graders (younger group, transitioning to complex math concepts like fractions), 2 = eleventh graders (older group, handling advanced algebra and geometry). This variable distinguishes the two developmental cohorts in the study.<b>AMP</b>: Raw score on the Applied Math Problem-Solving subtest from the KeyMath-3 Diagnostic Assessment (translated to Hebrew). Represents the number of correct answers on contextual math problems involving algebra, geometry, measurement, and data analysis. Higher scores indicate better performance in interpreting and solving real-life math scenarios. Maximum possible varies by age-appropriate items; no time limit, calculators allowed.<b>RC</b>: Raw number of correct answers on the age-appropriate Reading Comprehension test. For fourth graders (Group 1): Based on the text "The Magic of Dragonflies" with multiple-choice questions (up to 32 items, evaluating literal, interpretive, and evaluative comprehension). For eleventh graders (Group 2): From a previous Israeli Psychometric Entrance Test passage (up to 13 items, assessing passage understanding, inference, syntax, and logical connections). Higher scores indicate stronger comprehension of informative texts.<b>MathFluency</b>: Raw score on the Arithmetic Fluency subtest from the Woodcock-Johnson IV Tests of Achievement. Represents the number of correct responses to simple arithmetic operations (addition, subtraction, multiplication) within a 3-minute time limit. Higher scores reflect greater fluency in basic computations, which partially mediates the RC-AMP relationship in younger students per the study.<b>ReadingFluency_Accuracy</b>: Number of errors made in the single-word reading subtest from the "Alef-Ad-Taf" Hebrew reading assessment (36 context-free words read aloud). Lower scores (fewer errors) indicate higher accuracy in decoding Hebrew words, which includes morphological complexity and vowel diacritics (nikkud). This measure contributes to overall reading fluency, with developmental shifts noted (speed-based in younger students, accuracy-driven in older ones).<b>ReadingFluency_Speed</b>: Time in seconds to complete the single-word reading subtest (36 words). Lower scores (faster times) indicate higher reading speed. This assesses decoding efficiency in Hebrew, influenced by orthographic features like transparent letter-sound mappings and root-based structures.<b>WM</b>: Raw score on the Backward Digit Span subtest from the Wechsler Intelligence Scale for Children. Represents the sum of correct trials where participants repeat number sequences in reverse order (up to 8 items, 2 trials each; 1 point per correct trial). Higher scores indicate better working memory capacity for holding and manipulating information, which correlates with arithmetic fluency in both groups but diminishes in influence for AMP and RC in older students.<b>RC per_corr</b>: Percentage of correct answers on the Reading Comprehension test, calculated as (RC / total items) * 100. For fourth graders: Based on up to 32 items (e.g., 10/32 = 31.25%). For eleventh graders: Based on up to 13 items (e.g., 11/13 ≈ 84.62%). This is the primary RC metric used in analyses, showing significant positive associations with AMP in both groups.For full methodological details, including test reliabilities (e.g., AMP: .89; WM: .93), administration procedures, and ethical approvals, refer to the "Materials and Methods" section of the manuscript. Data were analyzed using Spearman correlations and linear regressions due to non-normality. This dataset supports the proposed Developmental Linguistic-Cognitive Scaffold Model (DLCSM) in the manuscript. Note: Participant IDs are anonymized; no personally identifiable information is included.