AI-Enabled Special Education Services: The Moderating Role of Parental Involvement in Home-School-Community Collaboration

The implementation of inclusive education, a cornerstone of equitable learning opportunities for children with special needs, is increasingly reliant on a complex interplay of institutional support and familial engagement. A recent study published in Frontiers in Psychology sheds critical light on this dynamic, revealing that the perceived effectiveness of Artificial Intelligence (AI)-enabled special education services is significantly amplified by the depth and nature of parental involvement. While AI technologies offer promising avenues for personalized learning and support, their true potential in enhancing educational outcomes for children with special needs is intrinsically linked to the active participation of parents, according to the research.

Understanding the Landscape of Inclusive Education and AI Integration

Inclusive education, as defined by UNESCO, aims to create learning environments where all children, regardless of their abilities, can access quality education alongside their peers in mainstream settings. This educational philosophy has gained considerable traction globally, with governments increasingly embedding it within legal frameworks and policy initiatives, recognizing it as an educational right. However, the path to successful implementation is fraught with challenges, ranging from resource allocation to pedagogical adaptations.

In recent years, Artificial Intelligence has emerged as a transformative force in special education. AI-driven personalized learning platforms, sophisticated communication aids, and intelligent assessment tools are revolutionizing how special education is delivered. Studies have indicated that these digital assistive technologies hold immense potential for improving learning outcomes and enhancing the overall well-being of disabled learners. For instance, research by Hussein et al. (2025) highlights the pervasive use of AI in delivering tailored educational experiences, while Pang and Datu (2025) underscore the positive impact of these technologies on learning and well-being.

The Crucial Role of Parental Engagement

Despite the technological advancements, the effectiveness of special education services is not solely dictated by the sophistication of the tools employed. The attitudes, beliefs, and knowledge of parents play a pivotal role. A meta-analysis by Goldman and Burke (2017) underscored parental involvement as a key determinant in improving the effectiveness of special education. When it comes to AI-supported special education, a parent’s perception of the technology’s usefulness directly influences their willingness and level of engagement.

Epstein’s (2018) framework for home-school-community partnerships categorizes effective parental involvement into multifaceted components, including home learning support, home-school communication, and utilization of community resources. These elements collectively contribute to the overall efficacy of educational service delivery. However, the existing literature has often examined AI services and parental engagement in isolation, neglecting to explore how the level of parental engagement might moderate the perceived effectiveness of AI-driven interventions. While some studies, like that of Piccolo et al. (2024), have suggested parental involvement as a potential condition for the success of robot-mediated interventions, empirical evidence confirming this role has been limited. Similarly, Jang’s (2023) meta-analysis confirmed the effectiveness of parent education programs for children with disabilities but did not delve into the interaction between parental involvement and technology-specific factors. Guo and Keles (2025) found that intervention effectiveness varied with parental engagement types and intensity, but did not model parental involvement as a moderating condition. This gap highlights the need for more granular research into the moderating function of parental involvement in AI-assisted special education.

Theoretical Frameworks Guiding the Research

To understand the combined influence of AI empowerment and parental engagement on perceived service effectiveness, the study draws upon two key theoretical frameworks: the Technology Acceptance Model (TAM) and self-efficacy theory.

The TAM, developed by Davis (1989), posits that perceived usefulness and perceived ease of use are primary drivers of technology adoption. This model has been widely applied in educational contexts to explain how users, including parents, accept AI support in special education.

Self-efficacy theory, as articulated by Bandura (1997), provides a psychological lens for understanding parental involvement. This theory identifies four sources of efficacy judgments: mastery experiences, vicarious experiences, social persuasion, and physiological states. Mastery experiences, gained through direct success, are considered the most influential. Research has consistently shown a strong positive link between self-efficacy and positive attitudes towards inclusive education, suggesting that individuals with higher self-efficacy are better equipped to support others. Altindağ Kumaş and Sardohan Yildirim (2024) found that parents with higher self-efficacy are more adept at utilizing technological tools for their children’s education.

Parental attitudes, a critical factor influencing perceived service effectiveness, encompass affective (feelings about inclusive education), cognitive (beliefs about AI service efficiency), and behavioral (willingness to engage) components. Dignath et al. (2022) explored the formation of teachers’ beliefs about inclusive education, and their findings on attitude formation processes offer a parallel framework for understanding parental attitudes. The concept of family quality of life (Summers et al., 2005) further emphasizes satisfaction with services and perceived family efficacy as key elements in special education service effectiveness, closely linked to technology acceptance and family well-being.

In this study, TAM provides the theoretical basis for understanding parents’ evaluations of AI-enabled services through perceived usefulness and ease of use. Self-efficacy theory offers insights into why active parental involvement, particularly through home learning support, might amplify the association between AI empowerment and perceived effectiveness. Parental involvement thus acts as a boundary condition (Hayes, 2018) that shapes the strength of the technology perception-effectiveness link. Different forms of involvement draw on distinct efficacy sources: vicarious learning for home-school communication and mastery experience for home learning support (Bandura, 1997). While self-efficacy and parental attitudes are invoked as interpretive frameworks, they were not directly measured as constructs in the empirical model.

AI Applications and Parental Participation in the Chinese Context

China has made significant strides in special education policy over the past decade, with a growing emphasis on inclusive education within a rights-based continuum (Qu, 2024). The concept of "appropriate inclusion," localized to the Chinese socio-cultural context, has been proposed (Shen and Yin, 2025). Despite policy advancements, practical implementation faces structural barriers and unequal resource distribution (Alduais et al., 2023). In 2023, over 912,000 students with disabilities were enrolled in special education programs, with a substantial majority placed in mainstream settings through inclusive education. However, access to technology-assisted instruction remains uneven, particularly in rural and economically underdeveloped areas.

AI technologies are rapidly expanding in China’s special education sector, encompassing personalized learning tools, communication assistance, and intelligent assessment instruments (Hariyanto et al., 2025; Mukhtarkyzy et al., 2025). These interventions have shown demonstrable benefits for children with disabilities, particularly those with Autism Spectrum Disorder (Xu et al., 2026; Atturu et al., 2025).

Parental involvement in China exhibits culturally specific features and challenges. While Chinese parents generally hold favorable views toward educational involvement, a gap often exists between these views and actual engagement levels (Alduais et al., 2023). Cultural factors such as "face saving," a collectivist mindset, and an authoritative view of professionals can influence communication with school personnel. Resource constraints, exacerbated by the urban-rural divide, also pose significant challenges to facilitating parental involvement.

Existing research in China tends to study attitudes, efficacy beliefs, and participation experiences independently. The moderating role of parental participation in the efficacy-attitude dynamic remains underexplored. While meta-analyses confirm the effectiveness of parent education programs (Jang, 2023), they often fail to reveal how parental participation interacts with technological factors. This study aims to bridge this gap by investigating the moderating role of parental involvement in the relationship between AI-enabled special education services and perceived service effectiveness.

The Moderating Role of Parental Involvement: Key Findings

A study involving 386 parents of children with special needs in Shandong Province explored the moderating role of parental involvement. The research employed structural equation modeling to analyze the data.

The findings revealed a significant positive association between AI empowerment and perceived service effectiveness, with a standardized path coefficient ($beta$) of 0.31. This indicates that parents’ positive perceptions of AI-enabled services are directly linked to their belief in the effectiveness of these services. Family learning support also showed a direct positive association ($beta$ = 0.18) with perceived service effectiveness, suggesting that active parental support in learning activities at home contributes to the perceived value of the educational interventions.

However, the direct effect of home-school communication on perceived service effectiveness was not statistically significant ($beta$ = 0.06), suggesting that mere communication alone, without deeper engagement, may not directly translate into perceived effectiveness.

Crucially, both forms of parental involvement significantly moderated the relationship between AI empowerment and perceived service effectiveness. This means that the impact of AI services on perceived effectiveness is not constant but rather depends on the level of parental engagement. At low levels of involvement, the effect of AI empowerment on perceived service effectiveness was modest ($beta$ = -0.17 to -0.19). In stark contrast, at high levels of parental involvement, the effect of AI empowerment dramatically increased, reaching a large magnitude ($beta$ = 0.43 to 0.45).

Further analysis differentiated the roles of the two types of parental involvement. Home-school communication functioned as a "pure moderator," meaning its primary influence was in amplifying the relationship between AI empowerment and perceived effectiveness, rather than having a direct impact on effectiveness itself. Family learning support, on the other hand, acted as a "quasi moderator," exhibiting both a direct positive effect on perceived service effectiveness and a significant moderating effect.

These findings underscore a critical insight: the perceived effectiveness of AI-enabled special education services is substantially weaker in the absence of meaningful family involvement. This suggests that special education programs should prioritize integrating technology deployment with robust family-centered capacity-building initiatives.

Methodology and Data Analysis

The study involved 386 parents of children with special needs in Shandong Province who were utilizing AI-assisted special education services. Data were collected through a questionnaire survey, and structural equation modeling (SEM) was employed for analysis. The scales used in the study were validated through a translation-back-translation process and psychometric analyses, including confirmatory factor analysis (CFA) and reliability tests.

Common method bias was assessed using Harman’s one-way factorial test and by adding a common method factor to the measurement model, with results indicating that this bias was unlikely to be a dominant concern. Normality assumptions for the variables were checked through skewness, kurtosis, and Q-Q plots, and Mardia’s multivariate normality test. Composite scores were calculated, and interaction terms were created for the moderating analysis. Robust estimates were obtained using maximum likelihood estimation with bootstrap standard errors and 95% confidence intervals. Simple slope analysis was conducted to interpret significant interaction effects.

Implications and Future Directions

The study’s findings carry significant implications for educators, policymakers, and parents. Firstly, they highlight that technological innovation in special education must be complemented by strong family engagement strategies. Simply deploying AI tools is insufficient; fostering active parental participation is key to unlocking their full potential.

For educational institutions, this suggests a need to move beyond traditional one-way communication with parents. Instead, efforts should focus on enhancing the quality and problem-solving aspects of home-school interactions, particularly when it comes to AI services. For family learning support, targeted interventions such as parental training programs, resource libraries, and joint learning models for parents and children could be highly effective in strengthening both direct and moderating effects.

From a policy perspective, the research emphasizes the importance of a holistic support system that integrates AI advancement with parental capacity building. This includes not only technological infrastructure but also accessible professional counseling, community rehabilitation centers, and educational facilities that support both children and their families. The "appropriate inclusion" framework for China should consider these factors, addressing regional disparities and diverse family needs in resource allocation and support system design.

The study acknowledges limitations, including the sample being drawn from parents already engaged with AI-assisted services in a specific region, which may limit generalizability. Future research could explore these dynamics in broader populations and investigate the underlying attitudinal mechanisms more directly through direct measurement of self-efficacy and other psychological constructs. Longitudinal studies could also provide deeper insights into the causal relationships and long-term impact of parental involvement on AI-enabled special education outcomes.

In conclusion, this research provides compelling evidence that parental involvement is not merely a supplementary factor but a critical moderator in the effectiveness of AI-enabled special education services. By fostering robust family engagement, educational systems can significantly enhance the positive impact of technology, paving the way for more equitable and effective inclusive education for all children.