Next Article in Journal
Developing Middle School Students’ Systems Thinking in Earth Science Through Dynamic Simulations: The Case of the Rock Cycle
Previous Article in Journal
Transformative Potential in Special Education: How Perceived Success, Training, Exposure, and Experience Contribute to Teacher Readiness for Inclusive Practice
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Education Sciences
Volume 15
Issue 11
10.3390/educsci15111478
education-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Balancing Affective Engagement and Cognitive Load in Generative-AI-Based Learning: Empathy, Immersion, and Emotional Design in Design Education

by
Wonsub Lee
1,
Sungbok Chang
2,* and
Jungho Suh
3,*
1
Department of Design and Arts Management, College of Design and Arts, Hongik University, Seoul 04066, Republic of Korea
2
Department of Digital Media Design, College of Arts, Cheongju University, Cheongju 28496, Republic of Korea
3
Department of Media Communication, College of Social Science, Gachon University, Seongnam 13120, Republic of Korea
*
Authors to whom correspondence should be addressed.
Educ. Sci. 2025, 15(11), 1478; https://doi.org/10.3390/educsci15111478
Submission received: 25 September 2025 / Revised: 27 October 2025 / Accepted: 29 October 2025 / Published: 3 November 2025
(This article belongs to the Topic AI Trends in Teacher and Student Training)
Browse Figures
Versions Notes

Abstract

As higher education undergoes rapid transformation driven by Artificial Intelligence (AI), the integration of Generative AI (GenAI) has become essential for preparing future-ready creative professionals. In this context, design education plays a leading role in exploring how GenAI can enhance students’ experiential learning. This study empirically examined how three experience dimensions—Educational, Entertainment, and Aesthetic—shape Empathy, Immersion, Satisfaction, and Learning Outcomes in a GenAI-based self-character workshop. A total of 185 design students participated, and the data were analyzed using Structural Equation Modeling (SEM). The results revealed that both Entertainment (β = 0.334, p < 0.001) and Aesthetic (β = 0.434, p < 0.001) experiences significantly and positively predicted Empathy and also increased Immersion (β = 0.215, p < 0.001; β = 0.154, p < 0.05). In contrast, Educational experience showed a non-significant or slightly negative effect. Furthermore, Empathy enhanced Immersion (β = 0.220, p < 0.01), Satisfaction (β = 0.173, p < 0.05), and Learning Outcomes (β = 0.305, p < 0.001). Immersion also improved Learning Outcomes (β = 0.253, p < 0.05) but slightly reduced short-term Satisfaction (β = −0.186, p < 0.05), indicating a cognitive-load trade-off between concentration and immediate enjoyment. These findings demonstrate that GenAI-based creative activities can effectively foster both emotional engagement and learning performance when instructional design minimizes unnecessary cognitive burden. The study contributes to understanding how emotionally meaningful and aesthetically engaging experiences can advance AI-integrated design education in the digital transformation era.

1. Introduction

The rapid advancement of Artificial Intelligence (AI) has transformed not only the ways in which knowledge is produced but also how creativity is cultivated in higher education. Within this context, Generative AI (GenAI)—including tools such as large language models and text-to-image generators—has become a central force driving innovation in creative education. By enabling learners to co-create visual, textual, and interactive artifacts with AI systems, GenAI opens a new paradigm of learning that blends automation, authorship, and emotional engagement. As design disciplines are inherently situated at the intersection of creativity and technology, design education serves as an ideal testbed for exploring how GenAI can support experiential learning and human–AI collaboration (Zawacki-Richter et al., 2019; Yang & Lim, 2024; Long & Magerko, 2020).
Recent studies emphasize that learning in the GenAI era is no longer confined to traditional knowledge transfer but is increasingly experience-driven, encompassing affective and aesthetic dimensions that influence motivation and learning outcomes (Pine & Gilmore, 1999; Makransky & Mayer, 2022; Wijesundara et al., 2022; Li et al., 2024). Building upon Pine and Gilmore’s Experience Economy framework (Pine & Gilmore, 1999), the current study focuses on three core experiential dimensions in AI-enhanced learning: Educational experience (perceived instructional value and knowledge gain), Entertainment experience (enjoyment and playfulness), and Aesthetic experience (perception of beauty and design coherence). These dimensions are considered essential in shaping the learner’s emotional and cognitive engagement during creative activities. While entertainment and aesthetic experiences are expected to enhance engagement and empathy, educational components—when overly formal—may impose cognitive load that inadvertently reduces short-term satisfaction (S. Kim et al., 2024).
Beyond the affective aspects, two psychological mechanisms—Empathy and Immersion—are central to understanding the learning process in AI-mediated creative contexts. Empathy enables learners to identify emotionally with their creations or peers, fostering connection and reflection, while immersion represents deep absorption or flow in the learning activity (Makransky & Mayer, 2022; Guo & Kim, 2023; S. Kim et al., 2024). In GenAI environments, these mechanisms are particularly relevant because learners act as co-authors rather than passive recipients. Through creating an AI-assisted self-character, students externalize their identities, negotiate meaning, and experience a unique blend of creativity and introspection. This generative authorship process differentiates GenAI learning from earlier VR- or HCI-based educational experiences, where interaction was largely predefined by system design (Wijesundara et al., 2022; Yang & Lim, 2024).
Although prior research has demonstrated positive relationships among entertainment, aesthetic experience, and engagement, relatively few empirical studies have examined how these experiential dimensions interact with empathy and immersion to influence satisfaction and learning outcomes in GenAI-based design education (Guo & Kim, 2023; Li et al., 2024; Koh et al., 2023). Moreover, educational experience—while traditionally assumed to be beneficial—has received limited scrutiny regarding its potential negative or neutral effects under cognitive strain (S. Kim et al., 2024). Addressing this gap is essential for understanding how to balance affective engagement and instructional rigor in AI-supported pedagogical models.
Therefore, this study aims to empirically investigate how the Educational, Entertainment, and Aesthetic experiences influence Empathy, Immersion, Satisfaction, and Learning Outcomes in a GenAI-based self-character workshop (GSCW) for undergraduate design students. Using Structural Equation Modeling (SEM), we test an integrated model that links experiential dimensions to affective and cognitive outcomes. Specifically, we hypothesize that entertainment and aesthetic experiences enhance empathy and immersion, while educational experience may show a weaker or even negative effect due to cognitive load. By clarifying these relationships, this research contributes to the growing body of work on AI-driven experiential learning, offering practical implications for optimizing design education in the age of generative creativity.

2. Theoretical Background

2.1. AI-Aided Question and Content Generation in Education

The recent evolution of artificial intelligence (AI) has brought remarkable transformation to higher education. With the emergence of platforms such as GPT-4o, Copilot, and Gemini 2.5, universities are increasingly adopting AI tools to support teaching, assessment, and personalized learning. These systems offer educators significant efficiency gains by generating large quantities of questions and instructional materials in a short time, reducing cognitive workload, and enabling scalable education models (Kıyak & Emekli, 2024; Cheung et al., 2023; Gökçearslan et al., 2024).
AI-aided question and content generation provides several pedagogical advantages. It allows for rapid production of standardized materials, promotes adaptive learning through customized feedback, and enhances learner engagement by offering instant responses and self-assessment opportunities (Laupichler et al., 2024; Bai & Wang, 2025; Pitts et al., 2025; Ryan & Deci, 2000). However, the quality and depth of AI-generated materials often fall short of those created by human experts. For instance, medical-education studies have shown that chatbot-generated questions lack the conceptual depth, clinical relevance, and motivational elements required for higher-order thinking (Laupichler et al., 2024).
These limitations highlight the necessity of AI literacy, defined as the ability to critically evaluate, refine, and contextualize AI outputs to align with educational objectives. Without sufficient AI literacy, instructors and learners risk overreliance on automated tools, potentially leading to superficial understanding and reduced critical-thinking skills (Pitts et al., 2025). Moreover, ethical concerns—such as algorithmic bias, misinformation, and plagiarism—raise important questions about the validity and integrity of AI-generated content (Pitts et al., 2025).
As shown in Table 1, AI-based question and content generation enhances efficiency and accessibility, yet simultaneously poses challenges related to cognitive depth, accuracy, and ethics. To maximize its educational benefits, future pedagogical frameworks should emphasize AI literacy, human–AI collaboration, and critical-evaluation mechanisms that ensure the quality and contextual relevance of AI-assisted learning.
In conclusion, AI-aided question and content generation offers a dual-edged impact in educational practice. It enhances efficiency, personalization, and engagement but remains constrained by issues of depth, reliability, and ethics. The cultivation of AI literacy—both among educators and learners—stands as a key prerequisite for harnessing these technologies responsibly and sustainably within higher education.

2.2. Educational, Entertainment, and Aesthetic Experiences

Learners’ experiences in technology-enhanced education have long been recognized as multidimensional constructs that influence affective and cognitive outcomes. Building upon Pine and Gilmore’s experiential framework (Pine & Gilmore, 1999), the Educational, Entertainment, and Aesthetic dimensions have been empirically validated in numerous learning contexts (Pine & Gilmore, 1999; Schmitt, 2011; Jin & Lee, 2021). Educational experience refers to the perceived instructional value and relevance of content, reflecting the learner’s cognitive engagement and perceived learning gain. Entertainment experience captures enjoyment, curiosity, and playful interaction that stimulate intrinsic motivation (Hamari et al., 2014). Aesthetic experience denotes the perception of harmony, beauty, and design coherence, which facilitates affective attachment and attentional focus during learning (Fayn & Silvia, 2019). Recent studies demonstrate that entertainment and aesthetic experiences are strongly associated with positive emotions, flow, and empathy, whereas excessive educational framing may induce cognitive strain or reduce perceived enjoyment (Y. S. Chang & Wu, 2023; H. Lee & Choi, 2020). Specifically, Huang and Hew (2022) found that in game-based and immersive learning settings, playful and visually pleasing experiences enhanced engagement more effectively than explicit instruction. Similarly, Y. Kim and Lim (2023) reported that aesthetic appeal positively influenced both immersion and satisfaction in design-related learning environments. These findings suggest that experiential richness—particularly entertainment and aesthetic value—can serve as critical predictors of empathy and immersion in GenAI-based design education.

2.3. Empathy and Immersion as Mediating Mechanisms

Empathy and immersion represent core psychological mechanisms linking experience to learning outcomes. Empathy, defined as the ability to understand and share others’ feelings, is crucial in creative and collaborative learning environments (Li et al., 2024; Batson, 2011). In digital media learning, empathy promotes perspective-taking and identification with one’s creative work, leading to greater engagement and deeper reflection (Green & Brock, 2000). Immersion, often conceptualized as the state of flow or deep involvement, reflects focused cognitive and emotional absorption in a learning activity (Makransky & Mayer, 2022; Csikszentmihalyi, 1990). Empirical studies have demonstrated the mediating role of empathy and immersion in AI-mediated or immersive learning. For example, J. Lee et al. (2022) found that empathy mediated the relationship between entertainment experience and satisfaction in virtual storytelling-based education. Likewise, Makransky and Mayer (2022) showed that immersion positively affected motivation and perceived learning through heightened presence in VR environments. Furthermore, cognitive load research indicates that immersion can sometimes lead to a trade-off—enhancing performance but temporarily lowering satisfaction due to mental effort (S. Kim et al., 2024; Mayer, 2022). Therefore, empathy and immersion jointly serve as affective and cognitive bridges between the learner’s experience and educational outcomes.

2.4. Satisfaction and Learning Outcomes

Educational satisfaction represents the learner’s immediate evaluation of an instructional experience, encompassing both enjoyment and perceived achievement (Oliver & DeSarbo, 1988). Learning outcomes, on the other hand, reflect the enduring cognitive or skill-based gains derived from that experience. Prior research consistently supports that empathy and immersion exert significant positive effects on satisfaction and learning performance (Guo & Kim, 2023; J. Lee et al., 2022; Choi et al., 2024). For instance, Guo and Kim (2023) verified that emotional engagement directly predicted satisfaction and indirectly improved learning performance in virtual courses. Similarly, S. Kim et al. (2024) found that immersion enhanced perceived competence and actual learning outcomes, although excessive immersion could slightly diminish short-term satisfaction. Collectively, these studies suggest a dynamic interplay: educational, entertainment, and aesthetic experiences stimulate empathy and immersion, which in turn increase satisfaction and learning outcomes. Yet, the balance between cognitive rigor and affective engagement remains crucial, particularly in GenAI-based design education, where cognitive load and creative autonomy must be carefully aligned to optimize learning effectiveness.

3. Research Model and Hypothesis Development

3.1. Research Model Overview

Based on the theoretical framework outlined in the previous section, this study proposes a structural model that explains how the three experiential dimensions—Educational, Entertainment, and Aesthetic experiences—influence learners’ Empathy, Immersion, Educational Satisfaction, and Learning Outcomes in a Generative AI-based self-character workshop (GSCW). The model reflects the assumption that affective and cognitive mediators (empathy and immersion) play crucial roles in transmitting the effects of experience to learning outcomes. Drawing from the Experience Economy Theory (Pine & Gilmore, 1999) and recent findings in immersive and AI-mediated education (Makransky & Mayer, 2022; Guo & Kim, 2023; S. Kim et al., 2024; Y. S. Chang & Wu, 2023; J. Lee et al., 2022), the model postulates that entertainment and aesthetic experiences, which embody enjoyment and visual engagement, are likely to elicit higher empathy and immersion. In contrast, educational experience—associated with instructional content and cognitive effort—may yield neutral or even negative effects due to potential cognitive overload (S. Kim et al., 2024; Mayer, 2022). Empathy is expected to foster immersion and satisfaction by facilitating emotional engagement and identity connection with the learning task (Li et al., 2024; Batson, 2011; J. Lee et al., 2022), while immersion enhances performance but may slightly lower immediate satisfaction, consistent with the cognitive-load trade-off effect (S. Kim et al., 2024; Mayer, 2022). The proposed conceptual model is illustrated in Figure 1, and the hypotheses are formulated accordingly.

3.2. Hypothesis Development

3.2.1. Effects of Educational, Entertainment, and Aesthetic Experiences on Empathy

Empirical research has demonstrated that entertainment and aesthetic experiences are positively associated with empathy and affective engagement in creative and immersive learning environments (Huang & Hew, 2022; Y. Kim & Lim, 2023; J. Lee et al., 2022). When learners find activities enjoyable and aesthetically appealing, they are more likely to identify emotionally with their outputs or peers, thereby heightening empathy (Li et al., 2024; Batson, 2011). Conversely, educational experience, which involves cognitive and instructional elements, may either support or hinder empathy depending on its cognitive demand (Y. S. Chang & Wu, 2023; Mayer, 2022). Hence, the following hypotheses are proposed:
H1-1. 
Educational experience positively influences Empathy.
H1-2. 
Entertainment experience positively influences Empathy.
H1-3. 
Aesthetic experience positively influences Empathy.

3.2.2. Effects of Educational, Entertainment, and Aesthetic Experiences on Immersion

Prior studies indicate that enjoyable and visually rich learning environments enhance immersion and flow (Makransky & Mayer, 2022; Hamari et al., 2014; Huang & Hew, 2022). Entertainment and aesthetic experiences encourage sustained attention and emotional absorption, enabling learners to become deeply engaged in the creative process (Fayn & Silvia, 2019; Y. Kim & Lim, 2023). In contrast, educational experience may not directly foster immersion and can even disrupt it when instructional demands are overly rigid (S. Kim et al., 2024; Mayer, 2022). Accordingly, the hypotheses are stated as:
H2-1. 
Educational experience positively influences Immersion.
H2-2. 
Entertainment experience positively influences Immersion.
H2-3. 
Aesthetic experience positively influences Immersion.

3.2.3. Effects of Empathy on Immersion, Satisfaction, and Learning Outcomes

Empathy promotes emotional connection and meaning-making in learning, which in turn facilitates immersion and satisfaction (Guo & Kim, 2023; Green & Brock, 2000; J. Lee et al., 2022). When learners empathize with their creations or peers, they are more likely to experience a sense of presence and involvement (Li et al., 2024; Csikszentmihalyi, 1990). Moreover, empathy can enhance perceived learning outcomes by motivating deeper reflection and sustained engagement (Batson, 2011; Choi et al., 2024). Therefore, we hypothesize
H3. 
Empathy positively influences Immersion.
H4. 
Empathy positively influences Educational Satisfaction.
H6. 
Empathy positively influences Learning Outcomes.

3.2.4. Effects of Immersion on Satisfaction and Learning Outcomes

Immersion, representing deep involvement and focus, is a critical determinant of learning success (Makransky & Mayer, 2022; S. Kim et al., 2024; Mayer, 2022). Studies have shown that immersive experiences lead to improved understanding, skill acquisition, and long-term memory (J. Lee et al., 2022; Choi et al., 2024). However, excessive cognitive engagement may increase mental fatigue, leading to a temporary reduction in short-term satisfaction—a phenomenon referred to as the immersion–satisfaction trade-off (S. Kim et al., 2024). Thus, the following hypotheses are proposed:
H5. 
Immersion positively influences Educational Satisfaction.
H7. 
Immersion positively influences Learning Outcomes.

3.2.5. Effect of Educational Satisfaction on Learning Outcomes

Educational satisfaction reflects both affective and cognitive evaluations of the learning experience. Prior empirical research supports its significant predictive role in academic achievement and learning persistence (Guo & Kim, 2023; S. Kim et al., 2024; Oliver & DeSarbo, 1988). When learners perceive the activity as enjoyable and meaningful, satisfaction reinforces motivation and confidence, leading to improved outcomes. Hence, the final hypothesis is
H8. 
Educational Satisfaction positively influences Learning Outcomes.

3.3. Conceptual Model

Figure 1 presents the proposed structural model, depicting the hypothesized relationships among variables. The model integrates emotional (Empathy), cognitive (Immersion), and evaluative (Satisfaction) mediators to explain how three experiential dimensions collectively shape Learning Outcomes in a GenAI-based educational context.

4. Method

4.1. Participants and Data Collection

This study involved 217 undergraduate design students enrolled in a course integrating AI-based creative learning at a Korean arts university. All participants took part in a Generative AI–based Self-Character Workshop (GSCW) aimed at exploring personal identity and creativity through AI tools. Participants ranged in age from 20 to 25 years (M = 22.1, SD = 1.9); 67% were female and 33% male. Approximately 39% of participants reported using AI tools (e.g., ChatGPT-4o, Midjourney, Stable Diffusion 3.5) on a daily basis. The workshop was conducted as part of the Digital Media Design course in a single 3-h session.
Participation was voluntary, and students received minor course credit for their involvement. Informed consent was obtained from all participants, and ethical approval was granted by the Institutional Review Board of G university (1044396-202303-HR-043-01). After the workshop, participants completed an online survey. Data-quality checks excluded responses with unrealistically short completion times or uniform answer patterns, resulting in a final sample of 185 valid cases.

4.2. Generative AI-Based Self-Character Workshop (GSCW) Protocol

The Generative AI-Based Self-Character Workshop (GSCW) proposed in this study consisted of six structured sessions designed to cultivate self-expression and creative exploration through generative AI tools. This process comprised three phases, and the results were derived as shown in Figure 2.
First. First, Prompt Ideation and Keyword Selection; Next, Image Prototyping and Iteration; and Finally, Presentation and Reflective Evaluation. Each stage provides a reproducible model for AI-mediated creative learning in design education.
  • Phase 1: Prompt Ideation and Keyword Selection
Students learned prompt engineering basics using freely accessible tools such as Bing Image Creator (DALL·E-based) and Stable Diffusion. Each participant selected identity-relevant keywords—e.g., age, look-alike animal, nickname, hobbies, profession, salient traits, and aesthetic preferences—and composed initial prompts. This phase served as an ice-breaking activity that expanded verbal self-introduction into visual self-expression, encouraging interaction and empathy.
  • Phase 2: Image Prototyping and Iteration
Using the initial prompts, participants generated prototype images, reviewed outputs, and iteratively refined the prompts (syntax, modifiers, style) to improve self-representation fidelity. Each student produced at least five prototypes, exploring variations in style, composition, and emotional tone. This iterative loop mirrors design thinking (ideation–prototype–test–refine) and helps students understand AI affordances and limits.
  • Phase 3: Selection, Sharing, and Reflection
Learners selected a final preferred image that best represented their self-character and presented it in a sharing session. They articulated creative rationales, aesthetic decisions, and emotional reactions, and engaged in peer feedback. This reflective stage reinforced social learning and helped students connect AI-generated artifacts with identity construction.
  • Duration and Instructional Framework
The workshop spanned six class sessions (2–3 h each). Day 1: generative AI foundations and self-expression; Days 2–3: prompt development and initial prototyping; Days 4–5: iteration, refinement, and critique; Day 6: final presentation and reflection. Formative feedback and technical guidance were provided throughout to align imagery with authentic self-expression.
  • Educational Significance
GSCW illustrates how generative AI functions as both creative medium and reflective mirror for self-exploration. Transforming self-introduction into visual co-creation helped students translate abstract self-concepts into concrete visual narratives, bridging emotional and cognitive engagement. The six-day structure offers a reproducible framework adaptable to diverse design and media education contexts.

4.3. Operational Definitions of Variables and Measurement Instruments

Table 2 provides the operational definitions for all latent variables used in this study. All constructs were assessed using validated multi-item scales (5-point Likert, 1 = strongly disagree to 5 = strongly agree). All reliability coefficients (α = 0.861–0.958; CR = 0.905–0.962) indicated strong internal consistency.

4.4. Validity and Reliability of Measurement Variables

To verify the validity of the constructs and measurement items established in this study, a confirmatory factor analysis (CFA) was conducted. During this process, items were refined to improve the overall model fit, and modification indices were applied to enhance goodness-of-fit. As a result, two items from Esthetics, one item from Immersion, and one item from Learning Outcomes were removed.
To analyze the validity of the measurement variables, both convergent validity and discriminant validity were examined. First, convergent validity was assessed using the Average Variance Extracted (AVE) and Composite Reliability (CR) values. As shown in Table 3, the AVE values ranged from 0.708 to 0.864 (≥0.50 acceptable), and the CR values ranged from 0.905 to 0.962 (≥0.70 acceptable), thereby confirming convergent validity. Next, discriminant validity was assessed by comparing the square roots of AVE with the inter-construct correlation coefficients.
As presented in Table 4, the highest correlation coefficient was between Educational Experience and Educational Satisfaction (0.627). Since the smallest square root of AVE (0.929) exceeded this value, discriminant validity was confirmed, and subsequent hypothesis testing was conducted based on the structural model.

4.5. Data Analysis and Model Evaluation

Data were analyzed using Structural Equation Modeling (SEM) in AMOS 28 following the two-step approach (Anderson & Gerbing, 1988). First, the measurement model was tested via CFA, as reported above. Second, the structural model was analyzed to test hypothesized relationships. To ensure robustness, bootstrap resampling (5000 iterations) was used to test indirect effects (Hayes, 2022). Common Method Bias (CMB) was assessed via Harman’s single-factor test; the first factor accounted for 32.5% of variance, below the 50% threshold (Podsakoff et al., 2012).

5. Result

5.1. Data Screening and Descriptive Statistics

After excluding incomplete or low-quality responses, 185 valid cases were retained for analysis. All variables satisfied the assumption of normality (|skewness| < 2, |kurtosis| < 7). Descriptive statistics showed that participants perceived the Generative AI-based self-character workshop positively, with moderately high means for experiential variables. No multicollinearity issues were found (VIFs < 3.0).

5.2. Measurement Model

A Confirmatory Factor Analysis (CFA) was conducted to verify construct validity. All standardized loadings exceeded 0.70, Average Variance Extracted (AVE) was above 0.60, and Composite Reliability (CR) exceeded 0.85. The model demonstrated good fit: χ2 = 489.976 (df = 324, p < 0.001), GFI = 0.845, AGFI = 0.806, RMR = 0.034, TLI = 0.952, CFI = 0.959, RMSEA = 0.053. All indices satisfied the recommended thresholds (CFI/TLI > 0.90, RMSEA < 0.08) (Kline, 2016).

5.3. Structural Model

The hypothesized model was tested using Structural Equation Modeling (SEM) in AMOS 28. The overall model fit was acceptable and consistent with the measurement model: χ2 = 548.476 (df = 330, p < 0.001), GFI = 0.831, AGFI = 0.792, RMR = 0.053, TLI = 0.938, CFI = 0.946, RMSEA = 0.060.

5.4. Mediation Analysis

Bootstrapping (5000 samples) following Hayes (2022) confirmed the mediation effects. Both Empathy and Immersion significantly mediated the influence of Entertainment and Aesthetic Experiences on Satisfaction and Learning Outcomes. Significant indirect effects were found for Entertainment → Empathy → Satisfaction (β = 0.058, 95% CI [0.019, 0.103]), Aesthetic → Empathy → Satisfaction (β = 0.075, 95% CI [0.032, 0.122]), Entertainment → Empathy → Learning Outcomes (β = 0.102, 95% CI [0.041, 0.164]), Aesthetic → Empathy → Learning Outcomes (β = 0.133, 95% CI [0.061, 0.205]), and Immersion → Satisfaction → Learning Outcomes (β = 0.067, 95% CI [0.022, 0.126]).
The findings indicate that Entertainment and Aesthetic Experiences significantly predict Empathy, and Table 5 further summarizes the path estimates, standard errors, critical ratios, and significance levels for all hypothesized relationships. Furthermore, this Empathy was found to positively influence Immersion, Satisfaction, and Learning Outcomes. The effect of Educational Experience on affective responses was not significant (p = 0.055 and 0.078), suggesting that structured cognitive engagement does not directly enhance emotional involvement. This partially supports cognitive load theory (Podsakoff et al., 2012) implying that instructional demand may reduce affective immersion in creative learning contexts. Notably, Immersion improved Learning Outcomes (β = 0.253, p = 0.028) but reduced Satisfaction (β = −0.186, p = 0.040), supporting the Immersion–Satisfaction Trade-off reported in prior studies (Makransky & Mayer, 2022). This indicates that highly immersive experiences increase cognitive effort, which may transiently reduce perceived enjoyment. Overall, the findings demonstrate that affective and cognitive mediators (Empathy and Immersion) play crucial roles in translating experiential learning into meaningful educational outcomes. Such results emphasize the necessity of balancing emotional stimulation and cognitive demand when integrating Generative AI into design education. Figure 3 illustrates the directional relationships among variables based on the significance of the results.

6. Discussion

6.1. Overview of Key Findings

The present study provides empirical evidence that Generative AI-based experiential learning significantly shapes students’ affective and cognitive engagement in design education. Entertainment and Aesthetic Experiences exert strong positive influences on Empathy, which subsequently enhances Immersion, Satisfaction, and Learning Outcomes. Conversely, Educational Experience demonstrated only marginal or non-significant effects, indicating that instructional or cognitively oriented engagement alone may not sufficiently evoke emotional connection. These results highlight the necessity of integrating affective and aesthetic dimensions into the design of creative AI-based learning environments (Csikszentmihalyi, 1990; Oatley, 2016). Moreover, the structural model confirmed a negative relationship between Immersion and Satisfaction (β = −0.186, p = 0.040), illustrating a cognitive–affective tension often observed in immersive learning contexts. Deep immersion increases performance and learning gains yet temporarily reduces perceived enjoyment due to heightened cognitive demand (S. Kim & Park, 2023; Sweller, 1988).

6.2. Affective Engagement as the Core Mechanism of Generative AI Learning

The strong association between Aesthetic and Entertainment Experiences and Empathy confirms that affective engagement plays a fundamental role in AI-based creative education. Aesthetic experience evokes emotional resonance and reflective appreciation, facilitating deeper involvement with design outcomes (Bruner, 1990; Leder et al., 2004). Similarly, entertainment-oriented activities promote intrinsic motivation and curiosity, which are essential for sustaining engagement in exploratory learning environments (S. Chang & Suh, 2025b). In the Generative AI-based workshop, students’ co-creation processes with systems such as ChatGPT and Midjourney allowed them to externalize self-expression and engage emotionally with AI-generated artifacts. This supports the notion of emotional flow—where learners experience both cognitive challenge and affective fulfillment—consistent with Csikszentmihalyi’s flow theory (Csikszentmihalyi, 1990; Csikszentmihalyi & Nakamura, 2010).

6.3. Empathy as a Mediating Mechanism

Empathy emerged as a key mediating factor linking experiential dimensions to learning outcomes. This aligns with prior research showing that empathy fosters emotional projection and reflective interpretation, both of which enhance creativity and comprehension (Batson, 2011; Lessiter et al., 2001). In AI-mediated contexts, empathy serves as a cognitive–affective bridge, enabling learners to interpret machine-generated outcomes through human-centered meaning-making. Empathetic engagement allows learners to internalize AI collaboration as a shared creative process rather than a purely technical one, thereby strengthening intrinsic motivation and conceptual understanding (Green & Brock, 2000; Plass et al., 2014). The mediation results further demonstrate that empathy amplifies both Immersion and Satisfaction, reinforcing its role as a central psychological mechanism in experiential and affective learning frameworks.

6.4. Cognitive Load and Instructional Limitations

The marginal influence of Educational Experience on Empathy (β = 0.149, p = 0.055) and Immersion (β = 0.114, p = 0.078) suggests that highly structured instructional approaches may impose excessive cognitive load, thereby reducing affective engagement (S. Kim et al., 2024; Sweller, 1988). This supports Cognitive Load Theory, which posits that emotional engagement declines when learners face competing cognitive demands. In AI-assisted creative learning, students must simultaneously manage algorithmic complexity and creative interpretation, resulting in increased cognitive effort that can overshadow emotional immersion. Hence, instructional design in AI-mediated education should balance structured guidance and exploratory freedom. Reducing extraneous cognitive load while promoting aesthetic reflection may allow learners to sustain motivation and achieve both emotional and intellectual satisfaction.

6.5. The Immersion–Satisfaction Trade-Off

A central contribution of this study is the identification of an Immersion–Satisfaction Trade-off, wherein deeper immersion enhances learning outcomes but reduces short-term satisfaction. This finding aligns with evidence from immersive and experiential learning research that describes a paradox between focus intensity and emotional pleasure (Csikszentmihalyi, 1990; Sweller, 1988). In the context of Generative AI learning, learners’ deep absorption in the co-creation process appears to demand considerable mental energy, leading to temporary fatigue but long-term educational benefits. Therefore, managing cognitive load and emotional pacing becomes essential in designing sustainable AI-based curricula. Educators should introduce reflection sessions and rest intervals to mitigate fatigue while maintaining immersion-driven learning effectiveness.

6.6. Educational Implications

The findings provide actionable insights for integrating Generative AI into design education. First, aesthetic and entertainment-based learning elements should be emphasized to strengthen empathy and motivation. Second, educators should balance cognitive challenge with emotional engagement through adaptive pacing and reflective discussion. Third, embedding AI tools as co-creative partners can enhance empathy-driven exploration while supporting self-expression. Finally, emotional design strategies—such as visual appeal, storytelling, and self-referential creation—should be systematically embedded to promote intrinsic motivation and deeper learning (Pine & Gilmore, 1999; Plass et al., 2014).

6.7. Theoretical Contributions

This study contributes to theoretical advancement in AI-enhanced experiential learning by integrating three key perspectives: flow theory, cognitive load theory, and emotional design theory. These perspectives jointly explain how empathy and aesthetic experiences mediate learning effectiveness in AI-mediated contexts (J. Lee et al., 2022; Sweller, 1988; Plass et al., 2014). By combining these perspectives, the study positions empathy and aesthetic appreciation as fundamental drivers of human–AI co-creativity and educational transformation.

7. Conclusions

This study explored how Generative AI-based experiential learning influences students’ emotional and cognitive engagement within design education. The findings demonstrate that Entertainment and Aesthetic Experiences meaningfully foster Empathy, which subsequently enhances Immersion, Satisfaction, and Learning Outcomes. In contrast, Educational Experience alone did not sufficiently evoke affective engagement, suggesting that emotional resonance and aesthetic meaning are critical drivers of learning effectiveness in AI-mediated creative environments (Csikszentmihalyi, 1990; Csikszentmihalyi & Nakamura, 2010).
A key contribution of this study is the identification of the Immersion–Satisfaction Trade-off: while deep immersion promotes learning performance, it can simultaneously reduce short-term enjoyment due to cognitive demand. This paradox reflects the tension between mental effort and emotional fulfillment commonly observed in immersive learning (S. Kim & Park, 2023; Sweller, 1988).
Such findings emphasize that learning design must balance cognitive challenge and emotional comfort to sustain long-term motivation and engagement. Empathy emerged as a crucial mediating mechanism, connecting aesthetic appreciation with reflective understanding—illustrating that emotional resonance plays a central role in human–AI co-creative learning (Plass et al., 2014).
From an educational perspective, Generative AI should not be viewed merely as a technological tool but as a co-creative partner that stimulates self-reflection, empathy, and emotional participation. Curricula that incorporate storytelling, aesthetic exploration, and emotional design strategies can enhance intrinsic motivation and deepen cognitive engagement. Practical implementations may include combining high-immersion tasks with emotional debriefing sessions, adjusting task complexity to reduce cognitive overload, and encouraging peer feedback and reflective journaling to reinforce empathy-driven learning. These approaches transform AI use from a technical activity into an emotionally meaningful educational experience, aligning creative exploration with emotional well-being.
Theoretically, this study contributes to the integration of Flow Theory, Cognitive Load Theory, and Emotional Design Frameworks into a cohesive understanding of affective–cognitive dynamics in AI-enhanced education (Csikszentmihalyi, 1990; Sweller, 1988; Plass et al., 2014). By positioning empathy and aesthetic experience as central constructs, the study advances a holistic model of learning that aligns technological innovation with human creativity and affective sustainability. Practically, the proposed framework offers guidance for designing emotionally engaging and cognitively sustainable AI-based curricula that support the digital transformation (DX) of higher education.
Despite its strong empirical foundation, this research acknowledges certain limitations. Reliance on self-report data may limit objectivity, and the sample—design students from a single institution—restricts generalizability. Future studies should adopt cross-cultural and longitudinal designs, combine objective performance indicators with behavioral or physiological data, and investigate how empathy and immersion evolve through repeated AI-based learning experiences. Such research can deepen our understanding of emotional sustainability and creative resilience in digital learning environments.
In conclusion, affective engagement is the cornerstone of meaningful learning in the era of Generative AI. Empathy and aesthetic experience transform learning from a process of information acquisition to one of emotional discovery and co-creation, reaffirming that the future of design education lies in the seamless synthesis of human emotion and machine intelligence—a foundation for DX sustainability and curriculum resilience in the digital transformation of higher education.

Author Contributions

Conceptualization, S.C.; Software, S.C.; Formal analysis, W.L.; Investigation, W.L.; Resources, W.L.; Data curation, W.L.; Writing—review & editing, S.C.; Visualization, W.L.; Supervision, J.S.; Project administration, J.S.; Funding acquisition, J.S. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Gachon University research fund of 2024 (GCU-202404830001).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of Gachon University (protocol code 1044396-202303-HR-043-01).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

All data generated or analyzed in the study are included in the article. However, not all of it has been made publicly available as it may compromise the privacy of participants. Further inquiries can be directed to the corresponding authors.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Admiraal, W., Huizenga, J., Akkerman, S., & Dam, G. T. (2011). The concept of flow in collaborative game-based learning. Computers in Human Behavior, 27, 1185–1194. [Google Scholar] [CrossRef]
  2. Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two-step approach. Psychological Bulletin, 103, 411–423. [Google Scholar] [CrossRef]
  3. Bai, Y., & Wang, S. (2025). Impact of generative AI interaction and output quality on university students’ learning outcomes. Scientific Reports, 15, 24054. [Google Scholar] [CrossRef]
  4. Batson, C. D. (2011). Empathy-induced altruism: A psychological approach. Current Directions in Psychological Science, 20, 169–173. [Google Scholar]
  5. Bruner, J. (1990). Acts of meaning. Harvard University Press. [Google Scholar]
  6. Chang, S., & Suh, J. (2025a). The impact of digital storytelling on presence, immersion, enjoyment, and continued usage intention in VR-based museum exhibitions. Sensors, 25, 2914. [Google Scholar] [CrossRef]
  7. Chang, S., & Suh, J. (2025b). The impact of VR exhibition experiences on presence, interaction, immersion, and satisfaction: Focusing on the experience economy theory (4Es). Systems, 13, 55. [Google Scholar] [CrossRef]
  8. Chang, Y. S., & Wu, B. (2023). Balancing fun and focus: The impact of entertainment on cognitive load in digital learning. British Journal of Educational Technology, 54, 265–280. [Google Scholar]
  9. Cheung, B. H. H., Lau, G. K. K., Wong, G. T. C., Lee, E. Y. P., Kulkarni, D., Seow, C. S., Wong, R., & Co, M. T. H. (2023). ChatGPT versus human in generating medical graduate exam multiple choice questions. PLoS ONE, 18(8), e0290691. [Google Scholar] [CrossRef]
  10. Choi, B., Kim, J., & Lee, S. (2024). Mediating effects of flow and empathy on satisfaction and learning performance in immersive VR learning. Computers & Education, 205, 104915. [Google Scholar]
  11. Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. Harper & Row. [Google Scholar]
  12. Csikszentmihalyi, M., & Nakamura, J. (2010). The concept of flow and its applications to learning and creativity. In Cambridge handbook of creativity (pp. 195–206). Cambridge University Press. [Google Scholar]
  13. Deci, E. L., & Ryan, R. M. (1985). Intrinsic motivation and self-determination in human behavior. Springer. [Google Scholar]
  14. Fayn, K., & Silvia, P. J. (2019). Beauty in aesthetics and learning: Emotion, engagement, and flow. Psychology of Aesthetics, Creativity, and the Arts, 13, 141–150. [Google Scholar]
  15. Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential of the concept, state of the evidence. Review of Educational Research, 74, 59–109. [Google Scholar] [CrossRef]
  16. Gillan, D. J., & Braden, D. A. (2016). Cognitive outcomes of interactive learning environments. Computers & Education, 98, 225–237. [Google Scholar]
  17. Gökçearslan, Ş., Tosun, C., & Erdemir, Z. G. (2024). Benefits, challenges, and methods of AI chatbots in education: A systematic review. International Journal of Technology in Education, 7, 19–39. [Google Scholar] [CrossRef]
  18. Green, M. C., & Brock, T. C. (2000). The role of transportation in the persuasiveness of narratives. Journal of Personality and Social Psychology, 79, 701–721. [Google Scholar] [CrossRef]
  19. Guo, J., & Kim, J. (2023). Emotional engagement in virtual learning environments: A structural equation modeling approach. Computers in Human Behavior, 138, 107418. [Google Scholar]
  20. Ha, S., Lee, J., & Suh, S. (2024). Measuring creative competence in AI-supported learning environments. British Journal of Educational Technology, 55, 1021–1037. [Google Scholar]
  21. Hamari, J., Koivisto, J., & Sarsa, H. (2014, January 6–9). Does gamification work? A literature review of empirical studies on gamification. 47th Hawaii International Conference on System Sciences (HICSS) (pp. 3025–3034), Waikoloa, HI, USA. [Google Scholar]
  22. Hayes, A. F. (2022). Introduction to mediation, moderation, and conditional process analysis (3rd ed.). Guilford Press. [Google Scholar]
  23. Herrera, F., Bailenson, J., Weisz, E., Ogle, E., & Zaki, J. (2018). Building long-term empathy: A large-scale comparison of traditional and virtual reality perspective-taking. PLoS ONE, 13, e0204494. [Google Scholar] [CrossRef]
  24. Huang, B., & Hew, K. F. (2022). Implementing gamified learning in design education: Effects on engagement and performance. Computers & Education, 181, 104459. [Google Scholar]
  25. Hudson, C. C., Matson, R. H., & Taylor, M. J. (2019). Understanding immersion in educational games. Computers in Human Behavior, 94, 214–223. [Google Scholar]
  26. Jennett, C., Cox, A. L., Cairns, P., Dhoparee, S., Epps, A., Tijs, T., & Walton, A. (2008). Measuring and defining the experience of immersion in games. International Journal of Human–Computer Studies, 66, 641–661. [Google Scholar] [CrossRef]
  27. Jin, S., & Lee, K. (2021). The experiential dimensions of digital learning: An integrative framework. Computers & Education, 168, 104212. [Google Scholar]
  28. Jo, H. (2022). Creative learning outcomes in digital storytelling-based education. Educational Media International, 59, 185–198. [Google Scholar]
  29. Kim, J. (2015). Effects of learner-centered strategies on satisfaction and learning outcomes in digital education. Korean Journal of Educational Technology, 31, 55–70. [Google Scholar]
  30. Kim, S., Kim, J., & Park, S. (2024). Cognitive load and affective experience in immersive learning: Mediating effects on satisfaction and performance. British Journal of Educational Technology, 55, 215–230. [Google Scholar]
  31. Kim, S., & Park, S. (2023). Understanding the emotional paradox in immersive learning. Computers & Education, 194, 104690. [Google Scholar]
  32. Kim, Y., & Lim, C. (2023). Aesthetic quality and learner engagement in design-based courses. International Journal of Art & Design Education, 42, 329–342. [Google Scholar]
  33. Kıyak, Y. S., & Emekli, E. (2024). ChatGPT prompts for generating multiple-choice questions in medical education and evidence on their validity: A literature review. Postgraduate Medical Journal 100, 858–865. [Google Scholar] [CrossRef] [PubMed]
  34. Kline, R. B. (2016). Principles and practice of structural equation modeling (4th ed.). The Guilford Press. [Google Scholar]
  35. Koh, J. H. L., Chai, C. S., & Lee, M. H. (2023). Technological pedagogical content knowledge (TPACK) and design thinking integration for AI-enhanced learning. Computers & Education, 196, 104698. [Google Scholar]
  36. Kolb, D. A. (1984). Experiential learning: Experience as the source of learning and development. Prentice Hall. [Google Scholar]
  37. Laupichler, M. C., Rother, J. F., Kadow, I. C. G., Ahmadi, S., & Raupach, T. (2024). Large language models in medical education: Comparing ChatGPT- to human-generated exam questions. Academic Medicine, 99(5), 508–512. [Google Scholar] [CrossRef]
  38. Leder, H., Belke, B., Oeberst, A., & Augustin, D. (2004). A model of aesthetic appreciation and aesthetic judgments. British Journal of Psychology, 95, 489–508. [Google Scholar] [CrossRef] [PubMed]
  39. Lee, H., & Choi, J. (2020). The dual impact of instructional and playful design on learner engagement. Educational Technology Research and Development, 68, 2029–2050. [Google Scholar]
  40. Lee, J., Park, H., & Shin, D. (2022). Emotional immersion and empathy in virtual storytelling-based education. Educational Media International, 59, 102–118. [Google Scholar]
  41. Lee, J. H., & Kim, M. J. (2019). A study on the factors affecting educational experiences in digital learning environments. Korean Journal of Educational Technology, 35, 95–110. [Google Scholar]
  42. Lessiter, J., Freeman, J., Keogh, E., & Davidoff, J. (2001). A cross-media presence questionnaire: The ITC-sense of presence inventory. Presence: Teleoperators and Virtual Environments, 10, 282–297. [Google Scholar] [CrossRef]
  43. Li, X., Chen, J., & Fu, H. (2024). The roles of empathy and motivation in creativity in design thinking. International Journal of Technology and Design Education, 34, 1305–1324. [Google Scholar] [CrossRef]
  44. Long, D., & Magerko, B. (2020, April 25–30). What is AI literacy? Competencies and design considerations. 2020 CHI Conference on Human Factors in Computing Systems (CHI ’20) (pp. 1–16), Honolulu, HI, USA. [Google Scholar]
  45. Makransky, G., & Mayer, R. E. (2022). Benefits of taking a virtual field trip in immersive virtual reality: Evidence for the immersion principle in multimedia learning. Educational Psychology Review, 34, 1779–1807. [Google Scholar] [CrossRef] [PubMed]
  46. Mayer, R. E. (2022). Cognitive load theory and digital learning design. Educational Psychologist, 57, 1–13. [Google Scholar]
  47. Norman, D. A. (2004). Emotional design: Why we love (or hate) everyday things. Basic Books. [Google Scholar]
  48. Oatley, K. (2016). Emotions and the story worlds of literature. Scientific Study of Literature, 6, 64–77. [Google Scholar]
  49. Oliver, R. L., & DeSarbo, W. S. (1988). Response determinants in satisfaction judgments. Journal of Consumer Research, 14, 495–507. [Google Scholar] [CrossRef]
  50. Pine, B. J., & Gilmore, J. H. (1999). The experience economy: Work is theater and every business a stage. Harvard Business School Press. [Google Scholar]
  51. Pitts, G., Marcus, V., & Motamedi, S. (2025). Ethical and social considerations of AI in education. AI Ethics in Education Review, 2, 33–47. [Google Scholar]
  52. Plass, J. L., Heidig, S., Hayward, E. O., Homer, B. D., & Um, E. (2014). Emotional design in multimedia learning: Effects of shape and color on affect and learning. Learning and Instruction, 29, 128–140. [Google Scholar] [CrossRef]
  53. Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. (2012). Sources of method bias in social science research and recommendations on how to control it. Annual Review of Psychology, 63, 539–569. [Google Scholar] [CrossRef]
  54. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation. American Psychologist, 55, 68–78. [Google Scholar] [CrossRef]
  55. Schmitt, B. (2011). Experiential marketing: How to get customers to sense, feel, think, act, relate. Journal of Marketing Management, 27, 777–782. [Google Scholar]
  56. Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12, 257–285. [Google Scholar] [CrossRef]
  57. Venkatesh, V. (2020). Determinants of perceived ease of use: Integrating control, intrinsic motivation, and emotion into the technology acceptance model. Information Systems Research, 11, 342–365. [Google Scholar] [CrossRef]
  58. Wijesundara, C. J., Chang, Y. S., & Lee, H. (2022). Interactive digital storytelling for education: A review of empirical studies. Computers & Education, 181, 104451. [Google Scholar]
  59. Woo, Y., & Hwang, W. Y. (2009). A study on learners’ satisfaction and motivation in web-based learning. Educational Technology Research, 25, 105–122. [Google Scholar]
  60. Yang, H., & Lim, J. (2024). Design education in the age of generative AI: Opportunities and challenges for creative pedagogy. Design Studies, 90, 102151. [Google Scholar]
  61. Yoon, S., Kim, S., & Choi, Y. (2011). Immersion and flow in interactive media education. Journal of Educational Media, 37, 125–138. [Google Scholar]
  62. Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education—Where are the educators? International Journal of Educational Technology in Higher Education, 16, 39. [Google Scholar] [CrossRef]
Education 15 01478 g001
Figure 1. Research model designed to examine the relationships between Experience Factors and Empathy, Immersion, Educational Satisfaction, and Learning Outcomes.
Figure 1. Research model designed to examine the relationships between Experience Factors and Empathy, Immersion, Educational Satisfaction, and Learning Outcomes.
Education 15 01478 g001
Education 15 01478 g002
Figure 2. Generative AI-Based Self-Character Outputs.
Figure 2. Generative AI-Based Self-Character Outputs.
Education 15 01478 g002
Education 15 01478 g003
Figure 3. Result of path estimation.
Figure 3. Result of path estimation.
Education 15 01478 g003
Table 1. Summary of Key Advantages and Limitations of AI-Aided Question and Content Generation.
Table 1. Summary of Key Advantages and Limitations of AI-Aided Question and Content Generation.
DimensionAdvantagesLimitations
EfficiencyRapid generation of test questions and materials, reducing instructors’ workload and preparation time (Kıyak & Emekli, 2024; Cheung et al., 2023).Requires expert review to ensure alignment with learning objectives (Gökçearslan et al., 2024).
Quality & ConsistencyProduces grammatically correct and standardized questions at scale (Laupichler et al., 2024).Often lacks conceptual depth and contextual relevance, especially in professional domains (e.g., medicine) (Laupichler et al., 2024).
PersonalizationEnables adaptive difficulty and individualized feedback to enhance engagement (Bai & Wang, 2025).Overreliance on automation may diminish learners’ critical thinking and creativity (Pitts et al., 2025).
Feedback &
Motivation		Provides immediate feedback, promoting self-directed learning (Ryan & Deci, 2000).May reduce intrinsic motivation when interaction feels artificial or mechanical (Fredricks et al., 2004).
Ethics & LiteracyExpands accessibility and supports inclusive learning environments (Admiraal et al., 2011).Risks include data bias, misinformation, and plagiarism; requires strong AI literacy (Pitts et al., 2025).
Table 2. Latent Variables, Operational Definitions.
Table 2. Latent Variables, Operational Definitions.
ConstructDefinition and SourceItemsExample ItemReliability
Educational
Experience		Perceived instructional value and learning relevance of the workshop (S. Chang & Suh, 2025a; J. H. Lee & Kim, 2019).4“This activity helped me understand creative learning more effectively.”α = 0.952/CR = 0.962
Entertainment ExperienceEnjoyment, playfulness, and fun associated with the workshop (Kolb, 1984; Deci & Ryan, 1985; S. Chang & Suh, 2025b).4“This activity was fun and engaging.”α = 0.910/CR = 0.922
Aesthetic
Experience		Perceived visual appeal, harmony, and design coherence (S. Chang & Suh, 2025b; Norman, 2004).3“This activity was visually pleasing and harmonious.”α = 0.861/CR = 0.905
EmpathyEmotional identification with one’s creation or peers (Lessiter et al., 2001; Herrera et al., 2018).5“I could emotionally connect with the character I created.”α = 0.909/CR = 0.924
ImmersionDeep absorption and involvement during learning (Hudson et al., 2019; Jennett et al., 2008; Yoon et al., 2011).3“I was deeply immersed in the activity and lost track of time.”α = 0.932/CR = 0.950
Educational
Satisfaction		Immediate satisfaction and enjoyment of the learning experience (Venkatesh, 2020; Woo & Hwang, 2009; J. Kim, 2015).4“I was satisfied with my learning experience in this workshop.”α = 0.956/CR = 0.960
Learning OutcomesPerceived improvement in creativity, reflection, and skill acquisition (Gillan & Braden, 2016; Jo, 2022; Ha et al., 2024).5“This activity improved my creative thinking and problem-solving abilities.”α = 0.958/CR = 0.962
Table 3. The result of confirmatory factor analysis.
Table 3. The result of confirmatory factor analysis.
PathsUnstandardized EstimateS.E.C.R.βAVECR
Educational
Experience		->EDU11--0.8670.8640.962
->EDU 21.0230.05917.2370.927
->EDU 31.0420.06815.4020.861
->EDU 40.9680.05816.6550.824
Entertainment Experience->ENT11--0.8470.7470.922
->ENT21.1020.07414.8370.880
->ENT31.0620.07414.2780.857
->ENT40.8370.07511.2090.727
Aesthetic
Experience		->AES21--0.7500.8290.905
->AES 31.2950.1379.4440.977
Empathy->EMP11--0.7930.7080.924
->EMP20.9740.08910.9180.754
->EMP31.0610.08412.6920.851
->EMP40.9570.08511.3250.777
->EMP50.9380.08111.5970.792
Immersion->IMM11--0.8950.8620.950
->IMM21.0150.06814.960.860
->IMM31.0130.07513.5340.804
Educational
Satisfaction		->SAT11--0.7560.8260.960
->SAT 21.1540.09412.2510.854
->SAT 31.1730.09312.6560.878
->SAT 41.2010.09213.0830.904
->SAT 51.0740.09711.1230.786
Learning
Outcomes		->LO21--0.8410.8360.962
->LO31.0370.1437.2390.880
->LO41.070.1487.250.883
->LO50.990.1466.7980.778
->LO61.0580.03828.1340.866
Table 4. The squared correlations and AVE of variables.
Table 4. The squared correlations and AVE of variables.
EducationEntertainmentAestheticEmpathyImmersionEducationSatisfaction
Educational Experience0.929
Entertainment
Experience		0.331 ***0.864
Aesthetic Experience0.347 ***0.440 ***0.910
Empathy0.372 ***0.571 ***0.585 ***0.841
Immersion0.362 ***0.554 ***0.493 ***0.582 ***0.928
Educational
Satisfaction		0.627 ***0.282 ***0.445 ***0.389 ***0.433 ***0.908
learning outcomes0.240 **0.1090.1190.0450.245 **0.330 ***0.915
The values on the diagonal represent the square root of the AVE (Average Variance Extracted) for each variable. The values below the diagonal represent the correlation coefficients. ** p < 0.01 and *** p < 0.001.
Table 5. Summary of the path estimates, standard errors, critical ratios, and significance levels for all hypothesized relationships.
Table 5. Summary of the path estimates, standard errors, critical ratios, and significance levels for all hypothesized relationships.
HypothesesPathsEstimateS.E.C.R.p-ValueSupported
H1H1-1Educational
Experience		→Empathy0.1490.0781.9210.055No
H1-2Entertainment
Experience		→0.3340.0734.610***Yes
H1-3Aesthetic
Experience		→0.4340.0884.957***Yes
H2H2-1Educational
Experience		→Immersion0.1140.0651.7600.078No
H2-2Entertainment
Experience		→0.2150.0643.378***Yes
H2-3Aesthetic
Experience		→0.1540.0742.0760.038Yes
H3Empathy→Immersion0.2200.0792.7940.005Yes
H4Empathy→Educational
Satisfaction		0.1730.0742.3560.018Yes
H5Immersion→Educational
Satisfaction		−0.1860.091−2.0500.040No
(Trade-off)
H6Empathy→Learning
Outcomes		0.3050.0923.311***Yes
H7Immersion→Learning
Outcomes		0.2530.1152.2030.028Yes
H8Education
satisfaction		→Learning
Outcomes		0.3560.1073.341***Yes
*** p > 0.001.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Lee, W.; Chang, S.; Suh, J. Balancing Affective Engagement and Cognitive Load in Generative-AI-Based Learning: Empathy, Immersion, and Emotional Design in Design Education. Educ. Sci. 2025, 15, 1478. https://doi.org/10.3390/educsci15111478

AMA Style

Lee W, Chang S, Suh J. Balancing Affective Engagement and Cognitive Load in Generative-AI-Based Learning: Empathy, Immersion, and Emotional Design in Design Education. Education Sciences. 2025; 15(11):1478. https://doi.org/10.3390/educsci15111478

Chicago/Turabian Style

Lee, Wonsub, Sungbok Chang, and Jungho Suh. 2025. "Balancing Affective Engagement and Cognitive Load in Generative-AI-Based Learning: Empathy, Immersion, and Emotional Design in Design Education" Education Sciences 15, no. 11: 1478. https://doi.org/10.3390/educsci15111478

APA Style

Lee, W., Chang, S., & Suh, J. (2025). Balancing Affective Engagement and Cognitive Load in Generative-AI-Based Learning: Empathy, Immersion, and Emotional Design in Design Education. Education Sciences, 15(11), 1478. https://doi.org/10.3390/educsci15111478

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop