Beyond Usefulness: How Social Connection Drives Student Acceptance of Learning Robots

Author: Denis Avetisyan

New research reveals that Chinese primary school students are more likely to embrace social robots for English language practice when they perceive them as warm and engaging companions, rather than simply helpful tools.

A mixed-methods study using the Computers Are Social Actors paradigm demonstrates that enjoyment and perceived social cues are key determinants of learner acceptance in educational robotics.

While technological advancements continually reshape educational tools, understanding learner acceptance remains a complex challenge. This study, ‘A sequential explanatory mixed-methods study on the acceptance of a social robot for EFL speaking practice among Chinese primary school students: Insights from the Computers Are Social Actors (CASA) paradigm’, investigated factors influencing young learners’ engagement with social robots for English language practice. Findings revealed that perceived enjoyment and social attributes-specifically warmth and anthropomorphism-were stronger predictors of acceptance than functional usefulness, suggesting that emotional connection is key. How can these insights inform the design of socially intelligent technologies that truly motivate and empower language learners?

The Challenge of Engagement: Cultivating Active Learners

Primary English language education frequently encounters a significant hurdle in sustaining student engagement, ultimately impacting the development of crucial communication skills. Conventional methods, often reliant on rote memorization and passive reception, can fail to capture the attention of young learners, leading to diminished motivation and reduced retention. This disengagement isn’t merely a matter of classroom management; it directly correlates with a slower acquisition of vocabulary, grammatical structures, and the ability to express thoughts coherently. The result is a potential gap in communicative competence that can hinder academic progress and limit future opportunities, as effective communication is foundational across all disciplines and essential for navigating a globalized world.

The difficulty many students face in mastering English stems, in part, from a pervasive lack of learning experiences tailored to their unique needs. Traditional classroom settings often employ a one-size-fits-all approach, neglecting the diverse ways individuals process and retain information. This absence of personalized practice means learners with varying strengths – whether visual, auditory, or kinesthetic – are frequently underserved. Consequently, students may struggle to apply grammatical rules or vocabulary in real-world contexts, leading to disengagement and hindering the development of fluent communication. Innovative approaches recognize that effective language acquisition demands opportunities for iterative practice, immediate feedback, and content that resonates with individual interests – elements often missing in conventional methods.

The pursuit of more engaging primary English language learning experiences increasingly centers on innovative technologies. Researchers are actively investigating applications like gamified learning platforms, augmented reality tools that immerse students in interactive scenarios, and personalized learning systems powered by artificial intelligence. These technologies aim to move beyond rote memorization and static textbooks, offering dynamic practice opportunities tailored to individual student needs and learning paces. The potential benefits extend beyond increased motivation; adaptive software can provide immediate feedback, identify areas where students struggle, and offer targeted support, ultimately fostering a more effective and enjoyable learning environment. Such tools promise to transform language acquisition from a passive reception of information to an active, personalized journey of discovery.

The ability to communicate effectively is no longer simply a desirable skill, but a foundational requirement for success in nearly every aspect of modern life. Equipping young learners with robust communication skills-encompassing not just grammar and vocabulary, but also critical thinking, active listening, and nuanced expression-directly impacts their future academic achievements, career opportunities, and civic engagement. Studies consistently demonstrate a strong correlation between early communication proficiency and long-term socioeconomic outcomes, highlighting the urgency of addressing engagement challenges in primary English language learning. Beyond professional advantages, effective communication fosters stronger personal relationships, promotes empathy, and empowers individuals to navigate an increasingly complex world with confidence and clarity. Therefore, prioritizing innovative approaches to cultivate these skills represents a vital investment in the next generation’s potential and overall societal well-being.

Lvbao: A Social Interface for Embodied Language Practice

Lvbao is a physical robot designed to assist Chinese primary school students in practicing spoken English. Unlike conventional language learning software or exclusively virtual tutors, Lvbao offers a tangible interface for interaction, aiming to increase engagement and provide a more immersive learning experience. The robot’s development specifically targets the needs of this demographic, recognizing potential anxieties surrounding English pronunciation and conversational skills. By providing a non-judgmental, embodied partner, Lvbao seeks to encourage more frequent and confident spoken practice, supplementing traditional classroom instruction and offering personalized support outside of scheduled lessons.

Lvbao’s design centers on establishing a social presence to reduce the anxiety often associated with practicing a foreign language, particularly for primary school students. This is achieved through non-verbal cues such as eye contact, head movements, and gestures, combined with a physically embodied form. Research suggests that this approach can lower affective filters – psychological barriers that inhibit language acquisition – and increase student willingness to participate. By fostering a sense of rapport and perceived social safety, Lvbao aims to motivate students to actively engage in spoken English practice, even when facing potential errors or communication challenges. The robot’s physical form is intended to create a more natural and comfortable interaction compared to purely digital interfaces.

The Lvbao robot employs a ‘Social Robot’ approach predicated on the principle that physical embodiment enhances human-robot interaction and learning outcomes. This design incorporates physical features – including movement, facial expressions, and vocalizations – to establish a perceived social presence. This presence is intended to foster a sense of connection with the student, reducing anxiety associated with language practice and encouraging more frequent and sustained participation. Unlike purely software-based solutions, Lvbao’s physical form aims to create a more engaging and motivating learning environment by capitalizing on innate human tendencies to respond to social cues and embodied interaction.

Traditional English language learning methods for Chinese primary school students often lack individualized attention and opportunities for immediate corrective feedback, hindering fluency development. Lvbao addresses these limitations through a system of personalized interaction and real-time support. The robot’s speech recognition and natural language processing capabilities enable it to analyze student responses, identify errors in pronunciation or grammar, and provide instant, context-specific feedback. This dynamic interaction contrasts with conventional classroom settings or rote memorization exercises, offering a learning experience tailored to each student’s proficiency level and pacing. Furthermore, Lvbao’s ability to sustain extended conversational exchanges provides increased practice opportunities, fostering confidence and improving spoken English skills beyond what is typically achievable in traditional learning environments.

The CASA Paradigm: Perceiving Robots as Social Actors

The Computers Are Social Actors (CASA) paradigm posits that humans inherently treat technological entities exhibiting anthropomorphic features-such as voice, facial expressions, or conversational ability-as if they were social beings. This interaction is not based on the actual intelligence or capabilities of the technology, but rather on deeply ingrained social cues and expectations. Consequently, individuals will often apply social rules – including politeness, reciprocity, and emotional attribution – when interacting with these technologies, influencing their behavior, perceptions of trustworthiness, and overall engagement. This instinctive application of social cognition extends to expectations regarding communication style, responsiveness, and even perceived intentionality, shaping the human-technology interaction in ways analogous to interpersonal relationships.

Student perceptions of the robot Lvbao were assessed across three dimensions critical to positive social interaction: perceived intelligence, perceived warmth, and perceived social presence. Perceived intelligence measured the degree to which students believed Lvbao possessed cognitive abilities, while perceived warmth evaluated its likability and approachability. Perceived social presence quantified the extent to which students felt Lvbao was genuinely ‘there’ and actively engaged in the interaction. These factors were evaluated independently to determine their relative contributions to overall student engagement and learning outcomes when interacting with the robotic tutor.

Perceived anthropomorphism, the extent to which students attribute human characteristics to the Lvbao robot, demonstrated a direct correlation with engagement and learning outcomes. Data analysis revealed that higher scores on the ‘Perceived Anthropomorphism’ scale corresponded with increased student participation in learning activities and a greater willingness to actively seek assistance from the robot. Conversely, lower anthropomorphism scores were associated with students viewing Lvbao solely as a tool, limiting their interactive behavior and potentially hindering knowledge acquisition. This suggests that the perception of human-like qualities is not merely a subjective response, but a significant factor influencing the effectiveness of the robot as a learning companion.

Student perceptions of a social robot, such as Lvbao, directly influence the learning experience by establishing the robot’s role beyond a simple tool. Positive perceptions – specifically regarding intelligence, warmth, and social presence – correlate with the robot being viewed as a helpful companion, fostering increased engagement and a more collaborative learning environment. Conversely, lacking these positive perceptions results in Lvbao being categorized as merely a technological device, potentially limiting student interaction and diminishing its effectiveness as a learning aid. This categorization impacts the level of trust and rapport established, fundamentally altering the nature of the human-robot interaction and subsequent learning outcomes.

Validating Effectiveness: Acceptance and Enjoyment as Key Indicators

Structural Equation Modelling (SEM) was employed to statistically validate the relationships between perceived social characteristics of the Lvbao robot and student acceptance, building upon the established Technology Acceptance Model (TAM). This methodology allowed for the simultaneous testing of multiple pathways and complex interactions between variables, including perceived usefulness, perceived ease of use, and perceived enjoyment, as they relate to behavioral intention. Specifically, SEM enabled the assessment of how various social characteristics of Lvbao influenced students’ perceptions of its usefulness and ease of use, ultimately impacting their intention to continue utilizing the robot for language learning. The model’s fit indices were evaluated to ensure a robust and reliable representation of the data, confirming the significant influence of these perceived social attributes on acceptance.

Thematic analysis of student interactions with Lvbao indicated a strong correlation between students’ perceptions of the system’s usefulness and ease of use, and their stated behavioral intention to continue using it. This analysis identified ‘Perceived Usefulness’ and ‘Perceived Ease of Use’ as significant predictors of continued engagement. Specifically, ease of use demonstrated a standardized path coefficient of 0.34 (p < 0.001) impacting behavioral intention, while perceived usefulness also contributed, though to a lesser extent, suggesting that students were more likely to continue using Lvbao if they found it both helpful and simple to operate.

Student interaction data indicated a high degree of perceived enjoyment while using Lvbao, suggesting its capacity to enhance learning motivation. Statistical analysis, utilizing Structural Equation Modelling, revealed a significant positive relationship between perceived enjoyment and behavioral intention, quantified by a standardized path coefficient of 0.27 (p < 0.001). This indicates that for every one standard deviation increase in perceived enjoyment, behavioral intention increased by 0.27 standard deviations, holding other variables constant. Notably, ease of use demonstrated a slightly stronger impact on behavioral intention, with a path coefficient of 0.34 (p < 0.001).

Empirical analysis using Structural Equation Modelling demonstrated a positive correlation between the introduction of social robots and student engagement in language learning. Specifically, perceived enjoyment and ease of use were identified as significantly stronger predictors of behavioral intention to continue using the Lvbao social robot than perceived usefulness. The standardized path coefficient for perceived enjoyment was 0.32 (p < 0.001), indicating a statistically significant impact on acceptance. This suggests that factors contributing to a positive user experience, such as enjoyment and usability, are more influential in driving adoption than simply perceiving the tool as useful.

The Future of Learning: Personalized Experiences and Adaptable Systems

Recent studies highlight the remarkable capacity of social robots, such as Lvbao, to transform educational practices by delivering uniquely personalized learning experiences. Lvbao achieves this through adaptive interactions, tailoring lessons and feedback to each student’s individual pace and learning style-a feat traditionally challenging in conventional classroom settings. Observations indicate that this personalized approach doesn’t merely impart knowledge, but actively cultivates heightened student engagement, evidenced by increased participation and demonstrable improvements in knowledge retention. The robot’s ability to provide consistent, patient support, coupled with its engaging interface, fosters a comfortable learning environment where students feel empowered to ask questions and explore concepts without fear of judgment, ultimately suggesting a promising future where technology and pedagogy converge to unlock each learner’s full potential.

Investigations into the enduring effects of social robot interaction on developing language capabilities represent a crucial next step for this emerging field. While initial studies demonstrate promising short-term gains in engagement and preliminary language skills, a comprehensive understanding of long-term cognitive and communicative development remains elusive. Researchers are poised to examine whether sustained interaction with social robots fosters deeper linguistic structures, improves pragmatic language use in real-world scenarios, and potentially mitigates communication challenges for individuals who might otherwise struggle with traditional learning methods. These longitudinal studies will need to carefully assess not only measurable language proficiency but also the subtle nuances of social communication, including emotional recognition and reciprocal interaction, to fully realize the potential benefits-and address any unforeseen consequences-of this innovative educational approach.

The potential for transforming education extends far beyond initial applications, as this technology promises to reshape learning across diverse age groups and disciplines. Currently focused on early language acquisition, the core principles of personalized interaction and adaptive learning can be readily applied to subjects like mathematics, science, and even the arts. Imagine immersive historical simulations guided by a social robot, or a personalized tutor providing tailored support in complex scientific concepts-the possibilities are vast. Furthermore, this technology isn’t limited to traditional classroom settings; it could empower remote learners, provide support for students with special needs, and offer lifelong learning opportunities for adults seeking to upskill or explore new interests. By fostering individualized learning paths and cultivating deeper engagement, expanding this technology’s reach represents a significant step toward a more accessible, effective, and enriching educational experience for all.

The convergence of robotics and educational theory is forging new avenues for student empowerment through highly adaptable learning experiences. This approach moves beyond standardized curricula, recognizing that each student possesses unique needs and paces of learning. By leveraging technologies like social robots, educators can cultivate environments where individualized attention isn’t a limitation, but a core feature. The result is not simply the delivery of information, but the fostering of genuine engagement and a deeper understanding of concepts. Ultimately, this personalized scaffolding aims to unlock each student’s inherent potential, building confidence and cultivating a lifelong love of learning that extends far beyond the classroom.

The study meticulously distills the complex interplay between technological utility and human perception. It reveals that acceptance isn’t merely a calculation of usefulness, but a response to warmth and perceived social presence – a concept elegantly foreshadowed by Paul Erdős, who once said, “A mathematician knows a lot, but the computer knows more.” This observation mirrors the findings; students aren’t evaluating the robot’s raw computational power for EFL practice, but responding to its social cues. The research emphasizes that enjoyment and anthropomorphism, not simply functional benefits, drive learner acceptance, aligning with a preference for elegance and inherent qualities over sheer complexity in any system – be it mathematical or educational.

Further Lines of Inquiry

The observed primacy of affective response over pragmatic evaluation deserves restatement. Utility, it appears, is frequently a post-hoc rationalization. Future work should systematically manipulate the degree of social cues – warmth, gaze, vocal prosody – to determine thresholds beyond which perceived usefulness becomes secondary, or even irrelevant. The current study illuminates how acceptance occurs, but offers little predictive power regarding when it will occur.

A limitation resides in the specific cohort. Generalizability to other age groups, or to learners outside of a Chinese cultural context, remains uncertain. A comparative analysis, examining acceptance across varying educational systems, might reveal culturally-specific moderators. Furthermore, longitudinal studies are needed to assess the sustainability of this initial acceptance, and its impact on actual language acquisition.

The field would benefit from a move beyond simply demonstrating acceptance. The relevant question is not whether these robots are liked, but whether they demonstrably improve learning outcomes-and at what cost. Focusing on quantifiable gains, rather than subjective impressions, would be a necessary, if less charming, progression.

Original article: https://arxiv.org/pdf/2604.12789.pdf

Contact the author: https://www.linkedin.com/in/avetisyan/