Scholarlink

<p>Research suggests that youth with attention-deficit/hyperactivity disorder (ADHD) are at elevated risk for chronic absenteeism from school, with long term consequences including educational underachievement, increased mental illness, and economic difficulties later in life. However, limited research has examined whether theoretical models of absenteeism risk derived largely from non-ADHD populations generalize to youth with ADHD, and data-driven methods to identify those at greatest risk remain underdeveloped. Machine learning approaches may complement traditional theory-driven analyses by evaluating whether established risk factors replicate in this population while accounting for complex interactions among predictors. Using data from a clinical sample of youth with ADHD, the current study applied machine learning methods to develop an algorithm capable of classifying youth with patterns of school absenteeism and identifying the most influential problem areas for classification. Parents/caregivers of 194 clinically-referred youth with ADHD (<i>M</i><sub>age</sub>: 10.7&#xa0;years; <i>SD</i><sub>age</sub> = 2.98; 21.1% girls, 77.3% boys, 1.5% transgender or other) reported on their child’s history of school absenteeism and also completed 80 items assessing bio-psycho-social-educational problem areas. Random forest analyses yielded a model that classified youth with versus without a history of school absenteeism with acceptable discriminability (AUCs = .75–.79). The most influential problem areas included interfering behavior at school and home, depressed mood or thoughts, inconsistent parenting, attachment disorder, and vulnerability to at-risk behavior. Results provided support for the use of actuarial classification approaches as a complement to theory-driven methods for identifying absenteeism risk in youth with ADHD and may inform early identification and targeted prevention efforts.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Machine Learning Approach to Classifying School Absenteeism Patterns in Youth with ADHD

  • 期刊论文

  • 发表日期：

    2026年6月25日

查看全文
  • Patrick K. Goh,
  • Maria A. Rogers,
  • Ashlyn A. W. W. A. Wong,
  • Grace S. Mellor,
  • Jess Whitley,
  • J. David Smith,
  • Natasha McBrearty,
  • Natasha Tatartcheff-Quesnel,
  • George J. DuPaul

摘要

Research suggests that youth with attention-deficit/hyperactivity disorder (ADHD) are at elevated risk for chronic absenteeism from school, with long term consequences including educational underachievement, increased mental illness, and economic difficulties later in life. However, limited research has examined whether theoretical models of absenteeism risk derived largely from non-ADHD populations generalize to youth with ADHD, and data-driven methods to identify those at greatest risk remain underdeveloped. Machine learning approaches may complement traditional theory-driven analyses by evaluating whether established risk factors replicate in this population while accounting for complex interactions among predictors. Using data from a clinical sample of youth with ADHD, the current study applied machine learning methods to develop an algorithm capable of classifying youth with patterns of school absenteeism and identifying the most influential problem areas for classification. Parents/caregivers of 194 clinically-referred youth with ADHD (Mage: 10.7 years; SDage = 2.98; 21.1% girls, 77.3% boys, 1.5% transgender or other) reported on their child’s history of school absenteeism and also completed 80 items assessing bio-psycho-social-educational problem areas. Random forest analyses yielded a model that classified youth with versus without a history of school absenteeism with acceptable discriminability (AUCs = .75–.79). The most influential problem areas included interfering behavior at school and home, depressed mood or thoughts, inconsistent parenting, attachment disorder, and vulnerability to at-risk behavior. Results provided support for the use of actuarial classification approaches as a complement to theory-driven methods for identifying absenteeism risk in youth with ADHD and may inform early identification and targeted prevention efforts.