Adult learners with attention-deficit/hyperactivity disorder (ADHD) face significant academic challenges due to inattention, hyperactivity, and impulsivity (Wolf & Wasserstein, 2006), along with executive function deficits (EFDs), which impair self-regulation, planning and task prioritization (Silverstein et al., 2020). Without intervention, these challenges can persist, and adult learners face academic challenges (Sibley & Yeguez, 2018). A promising solution is Self-Regulated Learning (SRL), which fosters metacognitive skills like planning and self-monitoring to help learners manage their academic tasks.
However, for adult learners with ADHD, implementing SRL strategies independently is challenging due to the underlying EFDs. Therefore, artificial intelligence (AI) offers promising support by recognizing and responding to user emotions (Glick et al., 2024), feedback (Fang et al., 2025), bit size evaluations (Huang & Hew, 2025), and scheduling daily notifications (Breitwieser et al., 2023). Several studies have been conducted on AI-driven tools that support SRL, such as the SRL chatbot (Ng et al., 2024), the Meta tutor system (Dever et al., 2023), and an emotional support robot (Fang et al., 2025). However, there is little research on how AI can provide additional support that meets the needs of adult learners with ADHD/EFDs.
This study is phase one of a multiphase project that will result in the development of an AI-driven application to assist adult learners with ADHD/EFDs in applying SRL strategies. To achieve this, it is important to explore how ADHD/EFDs impact SRL, and current AI usage for SRL support adults with ADHD use AI for SRL support. These insights will inform the development of AI tools for enhancing SRL. Therefore, this foundational study is guided by the following questions:
What phases of Self-regulated learning (forethought, performance, or reflection) are most impacted?
How do adults with ADHD currently use AI tools to support their SRL?
Participants included 12 adult volunteers with ADHD recruited through email, social media platforms (Facebook, LinkedIn, and EDUCAUSE). Participants were adults (18-65) who were currently enrolled in a university or vocational school within the last 12 months and had a clinical or self-diagnosis of ADHD. Including both clinically and self-diagnosed individuals provided a broader population, as ADHD can often go unrecognized, especially in children who are predominantly inattentive without disruptive behaviors. This oversight can persist into adulthood, leading to a significant number of undiagnosed cases (Asherson et al., 2012). Data was collected using a 29-item questionnaire with a combination of a 5-point Likert scale and choose all that apply, with space for additional written responses. The researcher-developed questionnaire focused on two main areas: 1) SRL strategies affected by ADHD/EFDs. 2) Participants' current use of AI to support SRL strategies on academic tasks. The questionnaire is based on Zimmerman’s SRL framework. Zimmerman’s SRL framework is a three-phase framework that includes a Forethought phase (goal setting, planning, and self-motivation), Performance phase (self-instruction, attention, and metacognitive monitoring), and Self-Reflection phase (self-adjustment and self-reflection). Quantitative data were analyzed using descriptive statistics, focusing on response frequencies.
To answer research question 1, we examined how ADHD/EFD impacts the three phases of SRL. Many participants found it easy to set academic goals but found difficulty following through with their goals (see Table 1). All of the participants had trouble managing their time while starting the task (see Table 2). The majority of participants had confidence in assessing their academic performance, but only sometimes adjusted their learning based on academic performance (see Table 3). Exploring the use of AI tools to help with academic assignments, more than half of the participants reported currently not using AI tools.
Table 1
Forethought Phase
Category | Response | Percentage |
Setting goals | Very difficult | 17% |
Somewhat easy | 42% | |
Easy | 25% | |
Confidence in ability to plan how to complete your assignments effectively | Definitely not confident | 17% |
Slightly Confident | 50% | |
Confident | 25% | |
Following through with specific goals | Very difficult | 33% |
Difficult | 33% | |
Somewhat easy | 33% |
Table 2
Performance Phase
Category | Response | Percentage |
Strategies used to plan academic tasks | Checklists | 67% |
Goal setting | 42% | |
Reminders/alarms | 33% | |
Do not use strategies | 17% | |
Following through with academic tasks | Very Difficult | 33% |
Difficult | 33% | |
Somewhat easy | 33% | |
Barriers following through with academic tasks | Trouble starting the task | 100% |
Mismanaged time | 100% | |
Forgot about set goals and assignments | 42% | |
Feeling like a fraud and don’t believe in own capabilities | 42% | |
Fear of not completing the task perfectly | 33% | |
Nothing prevents me | 8% | |
Other: lack of motivation & health issues | 8% | |
Other: I get distracted | 8% | |
Frequency of monitoring academic progress of task | Never | 33% |
Rarely | 25% | |
Sometimes | 33% | |
Often | 8% |
Table 3
Self-Reflection Phase
Category | Response | Percentage |
Confidence in the ability to assess academic performance | Definitely not confident | 8% |
Slightly confident | 33% | |
Confident | 50% | |
Very confident | 8% | |
Adjusting learning strategies based on academic performance | Never | 17% |
Sometimes | 67% | |
Often | 17% | |
Steps to evaluate academic performance | Reviewing what I did well | 50% |
Identify areas of improvement | 58% | |
Asking for feedback | 33% | |
Comparing performance to goals | 42% | |
I do not evaluate my performance | 25% |
Participants reported limited use of AI tools to assist with academic assignments. More than half of the participants did not use AI tools (58%), while one participant stated that they did not use AI because it could not solve their problems, and it was just a word generator. Meanwhile, ChatGPT (33%) and Grammarly (33%) were used by participants. The least popular tools used by participants were AI-based scheduling apps (17%) and AI-driven learning platforms (8%).
In this study, more than half of the participants found it somewhat easy or easy to set goals. However, there was a noticeable decline from goal setting to planning and finally to follow-through. This trend highlights a drop in success from initial goal setting to execution of goals. AI- driven tools such as pedagogical agents (Dever et al., 2023) could be used to prompt users to engage in SRL strategies. While a multimodal architecture (Huang & Hew, 2025) could identify needs and assist with goal setting and following through with the goals.
The results indicate that most struggle with time management and starting the task. Emotional barriers of perfectionism and lack of self-efficacy and confidence also hinder task initiation and completion. This signals a gap in the performance phase of SRL. However, AI-driven tools that recognize and respond to user emotions (Glick et al., 2024) in conjunction with intelligent feedback (Fang et al., 2025) can support users who struggle with perfectionism, lack of confidence, and self-efficacy. In addition, AI tools can assist with scheduling and daily notifications (Breitwieser et al., 2023) to support task initiation and time management.
The results from the self-reflection phase indicate that participants were moderately confident and engaged in evaluating their academic performance. This suggests this phase may be a strength for this population compared to the forethought and performance phases. However, the findings also highlight room for growth. While participants generally engaged in self-assessment, only a small number of participants reported sometimes adjusting their learning strategies based on past performance. This may indicate that while participants recognize areas for improvement, they struggle to translate these insights into actionable changes. Learner dashboards can support the transition from the Self-Reflection to Forethought phase, with users being able to view their academic patterns and outcomes.
Participants report limited use of AI tools to assist with their academic assignments. The stigma surrounding generative AI tools could contribute to the reluctance to use AI. Concerns about academic integrity and claims that using AI constitutes cheating, fosters laziness, or diminishes academic skills (Dempere et al., 2023) can further complicate students’ willingness to adopt AI tools.
This study contributes to theory and practice by examining the intersection of ADHD, EFDs, and SRL, along with AI, in the underexplored context of adult learners. While existing research focuses on children or general student populations, this study highlights the specific challenges and opportunities for using AI to support adult learners with ADHD. Future research will examine ADHD and motivation, conduct collaborative design sessions, and develop and test an AI-driven application prototype.
EdTech Archives