Early Identification of Children at Risk for Long-Term Difficulties with Mathematics Learning and Innumeracy in...

Area of Science:

• Educational Psychology
• Developmental Psychology
• Mathematics Education

Background:

• Students starting school with lower math skills often experience persistent difficulties throughout their academic careers and into adulthood.
• There is a lack of consensus on the most effective criteria for the early identification of students at risk for mathematics learning difficulties (MLD).
• Early identification is critical to implementing timely interventions and preventing long-term academic disadvantages.

Purpose of the Study:

• To assess the sensitivity and specificity of various criteria for identifying students at risk of MLD and adolescent innumeracy.
• To determine the optimal early identification criteria using a longitudinal study design.
• To analyze the academic growth trajectories of at-risk versus typically achieving students.

Main Methods:

• A longitudinal study tracked 445 students from 1st to 10th grade.
• Evaluated achievement-only criteria (math scores below 10th-35th percentiles in 1st grade and across 1st-2nd grade) and teacher-reported attentive behavior.
• Used latent score change (LCS) models to analyze mathematics growth from 1st to 10th grade, with 10th-grade innumeracy (<25th percentile) as the outcome criterion.

Main Results:

• An achievement cutoff below the 25th percentile in 1st grade demonstrated the best balance of sensitivity (.589) and specificity (.878) for identifying at-risk students.
• Latent score change models revealed that identified at-risk students began behind their peers in 1st grade, and this gap widened significantly over the school years.
• Not all students at risk were identified by the chosen criteria, suggesting that additional factors may need consideration in future research.

Conclusions:

• Low mathematics achievement in 1st grade, irrespective of cognitive ability, attentive behavior, or socioeconomic status, is a strong predictor of long-term MLD and adolescent innumeracy.
• The findings support the use of a 1st-grade mathematics achievement cutoff (below the 25th percentile) as a practical tool for early risk identification.
• Further research is needed to incorporate additional factors to improve the comprehensive identification of all students at risk for mathematics learning difficulties.