Enhancing Creative Self-Efficacy and Learning Motivation Through IRS-MFL and VPP Simulation in a Net-Zero Carbon Sustainability Course

Abstract

This study aims to evaluate how integrating an Interactive Response System (IRS) with Modified Flipped Learning (MFL) can enhance students’ attendance, learning motivation, and creative self-efficacy in a sustainability course contextualized by Virtual Power Plant (VPP) simulations. The IRS–MFL framework incorporated micro-learning videos, in-class real-time feedback, and collaborative learning activities to reduce cognitive load, alleviate learning anxiety, and promote engagement. Within this framework, the VPP simulations were not treated as an independent variable but rather as contextualized course content that provided authentic, sustainability-oriented problem scenarios for applying the IRS–MFL pedagogy. Quantitative analyses demonstrated significant improvements in attendance, reduced learning anxiety, and increased creative self-efficacy after the intervention. Qualitative reflections further revealed heightened motivation, deeper understanding, and strengthened systems thinking toward sustainability challenges. Collectively, these findings indicate that the IRS–MFL framework—contextualized through VPP simulations—effectively enhances both affective and cognitive learning outcomes, offering a replicable pedagogical model for sustainability-focused STEM education.

1. Introduction

Achieving sustainable development and addressing the grand challenges of the 21st century—such as climate change, energy transitions, and technological innovation—require a new generation of learners equipped with competencies that go beyond traditional subject knowledge. International frameworks, including the Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report and UNESCO’s global review of environmental education, have emphasized the urgency of embedding active, problem-based, and systems-oriented learning into curricula at all levels of education [1,2]. These calls underscore not only the importance of technical knowledge but also the need to cultivate higher-order skills that enable students to navigate complexity, uncertainty, and interdisciplinary problem-solving—skills that lie at the core of modern STEM education.

Low attendance, high learning anxiety, and weak self-efficacy have been increasingly recognized as persistent challenges in STEM and sustainability education. For instance, Doggrell [3] reported that in a medical laboratory science course, the average lecture attendance rate was relatively low, indicating a substantial decline in students’ physical participation even within professional science programs. Although no direct correlation between attendance and academic performance was found, the study revealed a broader pattern of disengagement in contemporary higher education, where students often substitute live lectures with digital or recorded materials.

In addition, Cooper et al. [4] demonstrated that large-enrollment active learning environments can inadvertently increase students’ anxiety when activities such as random cold calling or high-stakes group discussions are implemented without adequate support mechanisms. Their findings emphasize that the structure and delivery of interactive learning practices critically influence whether anxiety is heightened or mitigated.

Conversely, Ballen et al. [5] showed that enhancing students’ self-efficacy through inclusive active-learning environments significantly improved their academic performance and reduced achievement gaps among underrepresented learners. Collectively, these studies indicate that low attendance, elevated anxiety, and limited self-efficacy are interrelated barriers that undermine student engagement and achievement in STEM and sustainability-related contexts. Addressing these interconnected issues provides a strong pedagogical rationale for developing adaptive models—such as the IRS–MFL framework employed in this study—to simultaneously improve attendance, alleviate anxiety, and strengthen self-efficacy in net-zero carbon literacy education.

Despite growing consensus on the importance of such competencies, conventional STEM and engineering curricula often remain fragmented and overly theoretical. As a result, students are frequently offered limited opportunities to practice integrated competencies such as systems thinking, creativity, and decision-making under realistic constraints. In particular, systems thinking—the ability to identify causal feedback loops, interdependencies, and long-term consequences in complex socio-technical systems—is increasingly recognized as an essential cognitive skill in sustainability-oriented engineering education [6,7]. Without opportunities to develop such skills, students may gain declarative knowledge of sustainability standards but lack the capacity to apply this knowledge in authentic, dynamic contexts [8].

Within this context, sustainability-related energy systems offer a rich platform for authentic learning. Engineering design-based activities that simulate real-world energy environments can effectively foster integrated STEM competencies. Among these, Virtual Power Plant (VPP) simulations provide particularly valuable learning opportunities, as they mirror the complexity of modern energy systems by integrating generation, storage, demand response, and market coordination. These simulations require learners to interpret data, manage competing objectives, and evaluate trade-offs—thus reinforcing systems thinking, decision-making, and creative self-efficacy [9]. By situating abstract concepts within realistic and uncertain conditions, VPP-based activities help bridge the gap between theory and practice, offering students opportunities to engage with interdisciplinary and societally relevant challenges.

To maximize the educational potential of such simulation-based contexts, instructional strategies must be designed to lower learning anxiety, support differentiated preparation, and sustain student engagement. The integration of Modified Flipped Learning (MFL) and Immediate Response Systems (IRS) presents a promising approach in this regard. MFL, through concise and accessible pre-class video modules, allows students to grasp complex content at their own pace, while IRS tools provide real-time feedback and facilitate collaborative exploration during in-class sessions [10,11]. Together, these tools create a feedback-rich, inclusive learning environment that has been shown to enhance motivation, attention, and retention in STEM education [12].

Drawing upon these insights, this study implements an IRS–MFL instructional framework within a university-level STEM sustainability course, using VPP simulation as the central design-based activity. While prior studies have investigated flipped learning or simulation-based learning in isolation, there is a notable gap in research that integrates these technologies to evaluate their combined impact on student outcomes. Specifically, the synergistic effects of IRS-MFL and VPP simulation on key competencies like creative self-efficacy and learning motivation remain largely unexplored. This research seeks to fill this gap by providing empirical evidence on the role of this combined technology-enhanced pedagogy in preparing learners for real-world sustainability challenges.

Aligned with the United Nations Sustainable Development Goals (SDGs), this study not only demonstrates improved student engagement and learning outcomes but also contributes to the broader discourse on advancing STEM education for sustainability—an aim central to this Special Issue. Building upon this alignment, the course design explicitly contextualizes sustainability education within engineering-based learning frameworks. To clarify the scope of sustainability education, this course was explicitly aligned with three United Nations Sustainable Development Goals (SDGs): SDG 7 (Affordable and Clean Energy), SDG 13 (Climate Action), and SDG 4.7 (Education for Sustainable Development). These goals guided the curriculum design by integrating clean-energy engineering, carbon-neutral technologies, and systems-thinking skills into STEM education to foster net-zero carbon literacy.

1.1. Instructional Practice of IRS-MFL for Cultivating STEM Competencies

Integrating Modified Flipped Learning (MFL) with an Immediate Response System (IRS) offers a focused and responsive instructional strategy for addressing persistent learning challenges in technically demanding STEM domains—particularly those requiring complex and interdisciplinary problem-solving. Under this model, students first engage with structured pre-recorded videos and modular digital content that introduce foundational concepts, such as energy system operations, sustainability frameworks, and the dynamics of Virtual Power Plants (VPPs). These materials are intentionally concise and targeted, aiming to reduce cognitive overload and support individualized pacing.

Classroom time is then dedicated to active learning strategies, including collaborative simulations and scenario-based decision-making exercises that challenge students to apply theoretical concepts in realistic engineering and sustainability contexts [13,14]. These activities reinforce technical understanding while simultaneously promoting critical thinking, systems reasoning, and peer-based dialog—core competencies emphasized in effective STEM pedagogy.

To further enhance comprehension and engagement, the IRS platform is fully integrated into classroom activities. Instructors pose conceptual questions and simulation-driven prompts—such as optimizing VPP dispatch under fluctuating renewable output—to assess student understanding in real time. When misconceptions are identified, immediate clarification and structured peer discussion are initiated to strengthen conceptual clarity and reinforce systems-level reasoning. Prior research has shown that IRS tools, particularly in STEM settings, can reduce extraneous cognitive load while enhancing student attention, retention, and confidence [15,16].

This instructional design is grounded in constructivist and transformative learning theories, both of which emphasize learner agency, contextualized cognition, and iterative reflection [17,18]. Within this framework, students participate in activities such as modeling resource allocation strategies and evaluating multiple solution pathways under realistic technical and time-based constraints. These exercises are explicitly designed to cultivate systems thinking and foster an integrated understanding of complex data across spatial and temporal dimensions.

By combining structured digital preparation, interactive in-class application, and real-time feedback, the IRS–MFL approach creates a dynamic, student-centered learning environment that supports the development of essential STEM competencies. As further elaborated in Section 3.2, this instructional model not only reduces learner anxiety and disengagement but also enhances motivation, creative self-efficacy, and overall academic performance—outcomes that are explored in detail in the subsequent sections.

1.2. Enhancing Student Engagement and Confidence in STEM Learning via IRS-MFL

To frame the pedagogical rationale of this study, three interrelated challenges in STEM-based sustainability education were identified—low attendance, heightened learning anxiety, and weak self-efficacy. These factors have been commonly observed to undermine students’ engagement and performance, particularly in courses involving interdisciplinary and high cognitive-load content. Within the IRS–MFL instructional framework, these constructs are not treated as parallel outcomes but as interconnected mechanisms through which the intervention operates. Specifically, the model seeks to (1) enhance attendance through micro-learning and participation incentives, (2) alleviate learning anxiety via anonymous real-time feedback, and (3) cultivate creative confidence and applied problem-solving skills through contextualized and student-centered learning tasks. This holistic perspective provides a coherent foundation linking the pedagogical design to the learning outcomes examined in this study.

In this study, four key constructs—creativity, creative self-efficacy, learning motivation, and learning performance—were emphasized as the focal learning outcomes. Creativity is conceptualized as the learner’s ability to generate original, flexible, and feasible ideas when addressing complex sustainability problems. Creative self-efficacy reflects students’ confidence in their capacity to produce innovative and practical solutions, consistent with prior research on self-efficacy and creative performance [17,18,19,20]. Learning motivation follows the Self-Determination Theory framework, encompassing autonomy, competence, and relatedness as the core intrinsic factors driving student engagement and persistence [19,20]. Finally, learning performance denotes the observable integration of cognitive understanding and applied problem-solving competence within the IRS–MFL process. Clarifying these constructs reinforces the theoretical foundation of this study and provides a direct conceptual linkage between the literature review and the measurement instruments described in Section 2.4.

Improving student attendance, reducing learning-related anxiety, and fostering confidence in technical problem-solving are interrelated challenges in STEM-based sustainability education. These difficulties are particularly evident in courses such as carbon management, where students must navigate abstract concepts, interdisciplinary content, and demanding cognitive tasks—factors that often undermine engagement. To address these barriers, this study implemented a dual-instructional approach combining an Interactive Response System (IRS) with Modified Flipped Learning (MFL). The integrated model was designed to strengthen three dimensions of learning: (1) motivating class attendance, (2) reducing cognitive anxiety through real-time feedback, and (3) building creative self-efficacy in applying complex STEM concepts to authentic scenarios.

(1)

Motivation for Attendance

The MFL design enhanced student preparedness and increased the perceived value of in-class learning. Core topics—such as GHG categorization, ISO 14064-1:2018 [21] frameworks, and the operational logic of Virtual Power Plants (VPPs)—were introduced through concise pre-recorded modules. This self-paced format accommodated diverse learning styles and reduced the intimidation of technically dense material, allowing students to enter the classroom more confident and ready for application-oriented activities.

To sustain engagement, weekly participation credits and scenario-based team challenges were incorporated. These incentives aligned with self-determination theory, which emphasizes autonomy and competence as essential drivers of long-term academic motivation [19,20].

(2)

Reducing Cognitive Anxiety through Real-Time Feedback

The IRS was integrated into all classroom sessions to provide immediate, formative feedback and reduce anxiety when students encountered complex modeling or abstract logic. Real-time polls, conceptual checks, and simulations enabled instructors to detect misconceptions and clarify them instantly. For example, when students struggled to align renewable dispatch with 24/7 carbon-free electricity requirements, IRS-generated visualizations were used to illustrate temporal balancing step by step. This just-in-time feedback created a psychologically safe learning space, reducing the fear of error and fostering active participation [22,23].

Peer learning was reinforced by grouping students of mixed ability, while teaching assistants used IRS analytics to provide targeted scaffolding. This structure supported less confident learners, encouraged collaboration, and reduced competitive pressure. Optional after-class tutoring was offered to reinforce comprehension for students identified as at-risk, further alleviating lingering anxiety.

(3)

Building Creative Self-Efficacy and Active Problem-Solving

Beyond attendance and anxiety reduction, the IRS-MFL model was designed to cultivate students’ creative self-efficacy—their confidence in addressing novel and complex STEM challenges. Open-ended, scenario-based activities required integration of technical, environmental, and policy perspectives. For example, students designed decarbonization strategies for a university campus or simulated VPP operations under fluctuating renewable outputs. These tasks had multiple valid solutions, encouraging systems thinking, innovation, and reflective debate.

The IRS platform deepened this process by prompting reflective questions (e.g., “What trade-offs are involved in your dispatch strategy?” or “How would stricter carbon limits alter your solution?”), fostering metacognitive awareness and peer dialog [24,25]. Peer assessments, guided by rubrics emphasizing feasibility, carbon impact, and creative reasoning, further reinforced accountability and highlighted diverse approaches to problem-solving. This inclusivity encouraged broader recognition of student creativity and enhanced confidence in applying STEM knowledge [26].

The IRS-MFL framework functioned as a holistic instructional strategy to overcome persistent barriers in STEM-focused sustainability education. By combining flexible pre-class preparation, adaptive in-class interaction, and responsive post-class support, the approach improved attendance, reduced cognitive anxiety, and nurtured creativity, confidence, and systems thinking in the context of complex carbon-related challenges.

2. Materials and Methods

This study rigorously evaluated the effectiveness of the IRS–MFL instructional approach through a mixed-methods research design grounded in an instructional action research framework. The intervention took place in an undergraduate elective course, Net-Zero Carbon Management. Specifically, the study sought to determine whether this integrated pedagogical strategy could alleviate three persistent challenges in higher education: low student attendance, heightened learning anxiety, and inadequate academic performance.

This one-group pretest–posttest design was adopted to preliminarily examine the instructional effects of the IRS–MFL model under authentic classroom conditions where control-group allocation was not feasible. To support this design choice, Dimitrov and Rumrill [27] discussed the methodological practicality of one-group pretest–posttest designs in behavioral and educational research. Although such designs are susceptible to external factors (e.g., maturation effects), they remain suitable for preliminary evaluation of intervention outcomes, especially under real classroom conditions where resources or randomization are limited. The following subsections describe the course structure, instructional model, participant characteristics, data collection instruments, and analytical procedures employed to evaluate the intervention.

2.1. Instructional Planning and Course Development

In this study, the Virtual Power Plant (VPP) simulations were not treated as an independent variable of the instructional intervention. Instead, they served as contextualized course content aligning the IRS–MFL pedagogy with the learning objectives of the Net-Zero Carbon Sustainability curriculum. The simulations provided authentic sustainability scenarios for students to apply the IRS–MFL process, allowing the pedagogical mechanisms—rather than the simulations themselves—to influence the observed outcomes.

The IRS–MFL instructional model integrated multiple pedagogical components to address key challenges in sustainability-oriented STEM education.

• Micro-learning videos were employed to reduce cognitive load before class and enhance flexibility in learning access, thereby supporting consistent attendance.
• Anonymous IRS feedback provided real-time formative assessment and encouraged participation in a psychologically safe environment, mitigating students’ evaluation anxiety.
• Peer collaboration fostered social interaction and shared reflection, reinforcing engagement and persistence.

Finally, the Virtual Power Plant (VPP) simulations served as a practical illustration of the net-zero energy transition, allowing students to better understand the operation of sustainable energy systems and deepen their perception of the 2050 Net-Zero goal through systems-oriented thinking and contextual engagement. Collectively, these elements illustrated how the IRS–MFL framework, through contextualized learning scenarios and interactive mechanisms, improved student attendance, reduced learning anxiety, and enhanced creative self-efficacy without treating the VPP simulations as independent variables.

This study addressed persistent challenges in carbon management education—low attendance, elevated student anxiety, and limited academic achievement—by introducing a pedagogical model that integrates Interactive Response Systems (IRS) with Modified Flipped Learning (MFL). Implemented in the senior-level elective Net-Zero Carbon Management [28,29], the model was designed to strengthen both conceptual understanding and applied skills in GHG inventory methodology, ISO 14064-1:2018 categorization, Virtual Power Plant (VPP) operations, and 24/7 carbon-free electricity strategies.

Course development was guided by a four-phase action research cycle—planning, implementation, evaluation, and revision (Figure 1) [30]—which ensured continuous refinement of instructional practices and alignment with interdisciplinary objectives for net-zero carbon literacy.

During the planning phase, the instructional team undertook several preparatory activities to ensure both pedagogical rigor and practical relevance:

• Participated in targeted instructional design workshops to align with best practices in sustainability education;
• Consulted with external domain experts in carbon accounting, energy systems, and climate pedagogy;
• Co-developed interdisciplinary instructional content in collaboration with industry partners to ensure practical relevance;
• Analyzed student learning profiles and reviewed relevant literature to appropriately calibrate the course’s difficulty, pacing, and instructional tone.

To operationalize the course, a suite of preparatory materials and tools was constructed [31]. These included:

• Concise pre-class video modules (each 7–10 min), delivering essential theoretical content;
• Modular IRS question banks designed to assess both foundational and applied understanding;
• Collaborative simulation scenarios aimed at modeling real-world decarbonization decisions.

All instructional components were explicitly mapped to cognitive learning objectives (e.g., carbon reporting logic, systems integration) and affective outcomes (e.g., motivation enhancement, anxiety reduction). This strategic alignment aimed to create an engaging and psychologically supportive environment that promotes deeper conceptual engagement and sustained participation [32].

Furthermore, two core constructs were explicitly operationalized to align instruction and assessment. Creative Self-Efficacy was defined as students’ confidence in generating and applying innovative solutions to carbon-reduction and energy-system challenges, measured by a validated self-efficacy scale and supplemented by qualitative reflections from VPP project tasks. Learning Motivation was assessed through attendance records, voluntary engagement, and MSLQ-based survey items reflecting autonomy, competence, and intrinsic value under Self-Determination Theory (SDT). This explicit mapping clarified how affective and cognitive outcomes were integrated within the IRS–MFL framework.

2.2. Instructional Implementation and Model Differentiation

Building upon the course development framework, the implementation phase employed a blended IRS–MFL instructional model tailored to meet the interdisciplinary demands of net-zero carbon education. Foundational concepts were introduced through concise, pre-class videos, while in-class sessions emphasized active engagement via Virtual Power Plant (VPP) simulations, collaborative scenario-based problem solving, and IRS-facilitated formative assessments. Each instructional cycle followed a structured sequence [33]:

Real-time IRS responses enabled instructors to assess comprehension, dynamically adjust pacing, and clarify misconceptions during class. This responsiveness proved particularly beneficial for cognitively demanding topics such as ISO 14064-1:2018 scope distinctions and temporal-spatial carbon mapping. Furthermore, early detection of low-confidence learners through IRS analytics allowed for timely and personalized instructional support.

To reinforce conceptual understanding outside class, weekly tutorial sessions were facilitated by trained teaching assistants. These sessions provided additional guidance on simulation logic, dispatch modeling, and other abstract topics. Although voluntary, they were strongly recommended for students identified by IRS data as requiring supplementary instruction [34].

At the semester’s end, the instructional team conducted a comprehensive course review, synthesizing IRS interaction data, performance scores, post-course survey results, and student feedback from focus groups. Key improvements—such as refining video length, calibrating scenario complexity, and expanding the IRS item bank—were implemented to enhance future iterations.

To contextualize the instructional innovation, Figure 2 compares the designs of traditional lectures, conventional flipped classrooms, and the IRS–MFL model:

• Traditional Model: Students passively receive content during lectures and are expected to process and internalize material independently. This method often lacks timely feedback, contributing to low engagement in technically dense topics such as carbon accounting and system modeling [12].
• Conventional Flipped Model: Core content is delivered via long-form pre-class videos, freeing in-class time for interaction. However, students frequently report difficulty maintaining focus during extended video sessions, especially in STEM contexts involving abstract or quantitative content [35].

In contrast, the IRS–MFL model introduces two critical innovations:

• Microlearning Design: Pre-class videos were limited to 5–10 min, each focused on a single concept (e.g., ISO scope classification, VPP operational architecture). This design better aligned with cognitive load limitations and enabled targeted review of difficult concepts.
• In-Class IRS Integration: Real-time polling and adaptive feedback mechanisms empowered instructors to detect learning gaps and immediately reinforce or reteach essential concepts.

These features collectively fostered a high-engagement, student-centered environment. The model not only enhanced preparation and participation but also reduced perceived learning burdens, making it especially effective in high-load learning domains. It represents a refined application of flipped pedagogy designed specifically for sustainability-oriented STEM education.

Taiwan’s robust digital infrastructure further enabled successful implementation. The Taiwan Academic Network (TANet), established in the early 1980s, connects all universities and over 4000 schools with backbone bandwidths exceeding 10–100 Gbps. Additionally, the national broadband initiative—delivering fiber-to-the-home coverage to over 90% of the population—ensures equitable access across regions. Within this digital environment, IRS participation rates approached 100%, with rare disruptions attributed to individual device limitations. To ensure inclusivity, flexible response windows and low-tech alternatives (e.g., paper-based responses) were provided.

In summary, the IRS–MFL model demonstrated a scalable and theoretically grounded instructional strategy for advancing carbon literacy. By integrating simulation, formative assessment, and collaborative inquiry, it created an adaptive learning ecosystem that empowered students to engage deeply with the complex, interdependent challenges of net-zero carbon management. This study presents one of the first empirical applications of this pedagogical combination in sustainability education, offering a replicable model for STEM educators globally.

2.3. Participants

This study was conducted within a senior-level elective course titled Net-Zero Carbon Management, offered by the Department of Electrical Engineering at a private university in Taiwan. The course introduced students to key concepts in carbon accounting, greenhouse gas (GHG) inventory methodologies (in accordance with ISO 14064-1:2018), carbon footprint analysis, clean energy transition strategies, and the operational logic of Virtual Power Plants (VPPs) under 24/7 carbon-free energy procurement scenarios. To bridge theory and practice, the course adopted a blended instructional model combining lectures, case-based analysis, simulation-based exercises, and project-driven learning.

A total of 33 students enrolled in this 18-week course, comprising predominantly junior and senior undergraduates majoring in electrical engineering or related disciplines. Most participants had limited prior experience with carbon management or environmental accounting, and several expressed apprehension about the interdisciplinary and technical demands of the course. These characteristics created an authentic and challenging learning context—one well-suited for evaluating whether the IRS–MFL instructional model could effectively enhance student engagement and learning confidence in high-cognitive-load environments.

Classes met once a week for a 3 h session. While the primary language of instruction was Mandarin, bilingual materials were systematically employed to introduce internationally recognized standards and terminology. This ensured conceptual alignment with global carbon literacy frameworks and increased the accessibility of technical content. Throughout the semester, the IRS–MFL model was consistently applied to foster continuous feedback, reduce communication anxiety, and promote sustained student participation.

Participation in the research study was entirely voluntary and conducted in accordance with institutional ethical guidelines, as approved by the university’s Institutional Review Board (IRB). All students provided informed consent and were assured of data anonymity and confidentiality. To support authentic learning experiences, IRS polling activities and feedback loops were introduced as non-compulsory and formative in nature, minimizing evaluative pressure and reinforcing a learning-centered classroom culture.

By embedding this instructional innovation into a real-world classroom setting, the study aimed to assess how learner-centered, feedback-rich pedagogical strategies could address persistent challenges in sustainability education—namely, low engagement, heightened anxiety toward technical content, and difficulty transferring abstract knowledge to practical carbon management contexts.

2.4. Assessment and Data Collection

To comprehensively evaluate the effectiveness of the IRS–MFL instructional framework in addressing three primary learning challenges—(1) low attendance, (2) high learning anxiety, and (3) insufficient academic performance—this study employed a mixed-methods approach. Both quantitative and qualitative data were collected throughout the semester to assess levels of student engagement, affective development, and cognitive achievement.

2.4.1. Quantitative Instruments

Three instruments were employed to capture the quantitative aspects of student progress:

(a)

Attendance Records

Weekly attendance was tracked through IRS login and participation logs. This allowed for quantifiable measurement of student presence and engagement trends across different instructional sessions [36]. The collected data were compared with pre-IRS semester benchmarks to determine improvement in class participation.

(b)

Pre- and Post-Surveys

Structured surveys were administered at the beginning and end of the semester to assess shifts in students’ affective and cognitive states. The surveys included validated scales targeting four core constructs central to the IRS–MFL instructional objectives:

• Cognitive Anxiety toward technical topics—particularly in modeling abstract systems and understanding carbon dispatch logic—was measured using items adapted from the Cognitive Test Anxiety Scale (CTAS) by Cassady and Johnson, contextualized for STEM education [37].
• Creativity was assessed through the Creative Learning Environments Inventory (CLEI), focusing on students’ perceived ability to generate novel ideas in sustainability-related problem-solving [38].
• Creative Self-Efficacy was evaluated using Tierney and Farmer’s scale, which examines learners’ confidence in producing innovative and effective solutions [39].
• Learning Motivation was measured via select subscales from the Motivated Strategies for Learning Questionnaire (MSLQ), concentrating on intrinsic value and sustained engagement [40].
• Learning Performance was self-assessed through items adapted from prior studies on flipped and student-centered learning, capturing perceived comprehension and academic progress [41].

(c)

Learning Performance Metrics

Student performance was evaluated through scenario-based simulation tasks (e.g., VPP dispatch challenges) and selected IRS quizzes administered during class. These scores were normalized to allow cross-comparison of student gains, particularly in applied problem-solving contexts. Other project-based assignments were excluded from statistical analysis due to inconsistencies in completion and evaluation.

2.4.2. Qualitative Instruments

To supplement quantitative data, the study also incorporated qualitative methods to capture deeper insights into students’ experiences [42,43]:

(a)

Focus Group Interviews

Semi-structured interviews were held with six student groups (each consisting of 3–4 students) near the end of the course. These discussions explored perceptions of the IRS–MFL model, motivational changes, and instructional strengths and limitations. Themes related to anxiety reduction, confidence-building, and engagement were noted and triangulated with survey results.

(b)

Open-Ended Survey Responses

The post-course survey invited students to provide written reflections on how the IRS and MFL elements influenced their attention, comprehension, and classroom dynamics. Thematic coding was used to extract recurring perspectives and identify critical instructional factors.

(c)

Instructor Observation Logs

Throughout the semester, instructors maintained detailed logs documenting classroom interactions, student engagement, and real-time IRS feedback. These notes also captured affective observations such as increased student confidence, collaborative behaviors, and engagement fluctuations.

All qualitative data used in this study were collected exclusively from student participants. No instructor or teaching assistant perspectives were included in the qualitative dataset.

2.4.3. Data Analysis Strategy

Quantitative data were analyzed using paired-sample t-tests to assess pre–post changes in anxiety and motivation. Other statistical techniques such as regression or ANOVA were not used, ensuring alignment with the reported results. Qualitative data were examined using Braun and Clarke’s thematic analysis approach [44], allowing researchers to identify consistent patterns across interviews, surveys, and classroom observations.

This comprehensive multi-modal assessment framework enabled a robust evaluation of the IRS–MFL instructional model, capturing its influence on engagement, affective development, and academic performance in a complex sustainability learning context. Further elaboration on these results is provided in Section 3, supported by both statistical evidence and qualitative insights.

2.4.4. Software and Tools Description

The IRS used in this study was developed internally by Chung Yuan Christian University (CYCU, Taoyuan, Taiwan) as an in-house educational platform for real-time classroom interaction and feedback collection. In addition, Google Jamboard (Google LLC, Mountain View, CA, USA; service discontinued in 2024) and the iLMS system (Chung Yuan Christian University, Taoyuan, Taiwan), as well as Google Classroom (Google LLC, Mountain View, CA, USA; Available online: https://classroom.google.com/ (accessed on 14 September 2025)), were employed to facilitate student engagement, collaborative activities, attendance tracking, and course management. All software tools were used under institutional licenses provided by CYCU, ensuring data privacy and compliance with ethical guidelines.

3. Results and Discussion

This section presents the outcomes of implementing the IRS–MFL instructional model, focusing on its impact on student attendance, anxiety, and academic performance. The findings are organized into three subsections corresponding to the instructional challenges outlined earlier.

3.1. Improvements in Attendance and Engagement

The adoption of the IRS–MFL pedagogical framework led to notable gains in student attendance and engagement, as evidenced by improvements across five key behavioral metrics.

Table 1. Comparison of Attendance and Participation Metrics.

Indicator	Measurement Method	Pre-IRS (%)	Post-IRS (%)	Δ (%)
Formal attendance	IRS + manual verification	78	90	+12
Jamboard collaborative participation	In-class record	73	82	+9
Active listening/engagement rate	Instructor observation	63	70	+7
Post-class online survey participation	iLMS record	66	75	+9
Homework submission rate	iLMS/Google Classroom logs	69	80	+11

summarizes the comparative outcomes before and after the intervention, all of which exhibited statistically meaningful enhancements. Formal attendance, recorded via the IRS platform and verified manually, rose from 78% to 90%, reflecting a 12% increase in class participation. This improvement suggests a greater commitment to in-person learning sessions under the new instructional model.

To provide a comprehensive measure of attendance and engagement, this study defined Attendance as a composite indicator that reflects both physical presence and sustained participation within the IRS–MFL process. Consistent with Table 1, five complementary sub-indicators were used: (1) formal attendance recorded via the IRS platform and verified manually; (2) Jamboard collaborative participation based on in-class records; (3) active listening/engagement rate documented through instructor observation; (4) post-class online survey participation captured in the iLMS system; and (5) homework submission rate extracted from iLMS/Google Classroom logs. Each sub-indicator captures a distinct behavioral facet of engagement, and together they provide a holistic representation of learners’ in-semester participation. To maintain balance and avoid over-emphasizing any single dimension, the five sub-indicators were assigned equal weights (20% each) when computing the composite attendance score. This weighting scheme aligns with the participatory and feedback-oriented philosophy of the IRS–MFL framework, ensuring that the attendance construct reflects both quantitative presence and qualitative engagement.

Each of these sub-indicators captures a distinct behavioral aspect of student engagement, and together they provide a holistic representation of learners’ active involvement throughout the semester. To maintain balance and prevent overemphasis on any single dimension, the five sub-indicators were assigned equal weights (each contributing 20% to the composite attendance score). This weighting scheme aligns with the participatory and feedback-oriented philosophy of the IRS–MFL pedagogical framework, ensuring that attendance measurement reflects both quantitative presence and qualitative engagement.

In parallel, collaborative engagement—measured through active participation in in-class Jamboard exercises—rose by 9%, indicating increased willingness to engage in group-based scenario analysis and decision-making tasks. These quantitative improvements corroborate student testimonies gathered through post-course surveys and focus group interviews (see Section 2.4.2), where learners described feeling more motivated to participate due to the real-time feedback and interactive format enabled by the IRS.

Instructor-observed attentiveness, which includes indicators such as eye contact, note-taking, and responsiveness to prompts, improved from 63% to 70% (+7%). This result aligns with prior research suggesting that real-time polling systems can help maintain student focus in complex technical subjects by creating a more dynamic and responsive learning environment. Additionally, post-class online survey participation increased from 66% to 75%, a 9% gain that may reflect heightened student accountability and metacognitive engagement. The frequent use of IRS-based formative assessments and scenario competitions likely contributed to this outcome, as students were regularly prompted to reflect on their understanding and performance.

Among all metrics, the most significant improvement was observed in the homework submission rate, which rose from 69% to 80% (+11%). This suggests that the structured pre-class video modules in the flipped learning design not only facilitated better preparation but also helped reduce procrastination and assignment-related anxiety. Students noted in qualitative feedback that the bite-sized videos and clear instructional structure enhanced their confidence and ability to complete tasks on time.

Taken together, the consistent upward trend across all indicators underscores the efficacy of the IRS–MFL model in promoting student engagement. These findings can be interpreted through the lens of established educational theories. For instance, the improvements in attendance and homework submission rates directly support the tenets of Self-Determination Theory (SDT), which posits that learners are more motivated when their needs for autonomy, competence, and relatedness are met. The MFL component provides autonomy by allowing students to learn at their own pace, while the IRS fosters a sense of competence through timely feedback and boosts relatedness via collaborative activities. This alignment between our instructional design and SDT provides a theoretical basis for the observed behavioral changes.

3.2. Reducing Student Anxiety Through IRS-MFL

One of the central objectives of implementing the IRS–MFL pedagogy was to alleviate student anxiety—particularly in relation to abstract and technical topics such as carbon accounting frameworks, systems modeling, and dispatch simulation. To assess this impact, a structured anxiety inventory was administered at both the beginning and end of the semester, using a validated 9-item scale adapted from Cassady and Johnson [36]. Responses were recorded on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree), with high internal consistency (Cronbach’s α > 0.85).

As shown in

Table 2. Descriptive Statistics and t-test Results of Anxiety Inventory (N = 33).

No.	Statement	Pre-Test (Mean ± SD)	Post-Test (Mean ± SD)	t-Value
1	I find the technical terminology and concepts in this course difficult to understand.	4.42 $\pm$ 0.62	4.05 $\pm$ 0.73	7.62 ***
2	I feel most classmates understand this course better than I do.	4.30 $\pm$ 0.71	3.87 $\pm$ 0.83	7.43 ***
3	Repeated corrections from the instructor make me feel pressured when I still do not understand.	4.12 $\pm$ 0.69	3.69 $\pm$ 0.81	7.73 ***
4	Group projects, midterms, or final exams make me feel anxious or uneasy.	4.08 $\pm$ 0.74	3.68 $\pm$ 0.82	6.36 ***
5	I cannot fully understand the explanations provided by the instructor during this course.	4.16 $\pm$ 0.70	3.79 $\pm$ 0.82	6.87 ***
6	Watching pre-class video materials for this course causes me stress.	4.28 $\pm$ 0.62	3.85 $\pm$ 0.75	6.46 ***
7	I feel uneasy if the instructor’s grading criteria in this course are unclear or unfair.	4.36 $\pm$ 0.68	4.01 $\pm$ 0.78	7.65 ***
8	I worry about falling behind when the pace of this course is too fast.	4.32 $\pm$ 0.67	4.02 $\pm$ 0.77	7.25 ***
9	Missing post-class tutoring sessions makes me worry about keeping up.	4.09 $\pm$ 0.69	3.66 $\pm$ 0.81	5.35 ***

*** p < 0.001.

, statistically significant reductions were observed across all nine anxiety indicators. For instance, students’ initial difficulty in comprehending technical terminology was high (M = 4.42, SD = 0.62) but decreased notably following the IRS–MFL intervention (M = 4.05, SD = 0.73; t = 7.62, p < 0.001). Similarly, perceptions of pressure from instructor corrections (Item 3) and discomfort with the course’s pace (Item 8) both declined significantly (t = 7.73 and 7.25, respectively). These results suggest a broad reduction in both cognitive strain and emotional stress tied to course comprehension and classroom performance.

Several elements of the IRS–MFL model contributed to this outcome. The IRS component, used consistently throughout class sessions, provided immediate and anonymous formative feedback. This feature lowered the psychological risk associated with public errors. For example, when students struggled with the concept of temporal matching in 24/7 renewable dispatch scenarios, the instructor used real-time IRS polling and visual explanations to walk through the process step by step. This just-in-time intervention fostered psychological safety and helped correct misconceptions as they emerged.

The Modified Flipped Learning (MFL) strategy also played a key role. Foundational topics—such as GHG protocol scopes and VPP architecture—were delivered through short, focused pre-class videos that students could review at their own pace. This approach reduced cognitive overload and made complex terminology and processes more approachable. Compared to traditional full-length videos, these micro-learning modules preserved learner engagement without inducing fatigue, aligning with principles from cognitive load theory.

Moreover, the combination of in-class simulations and peer discussions, supported by IRS diagnostics, created a collaborative environment where uncertainty was normalized. During open-ended problem-solving tasks, students were encouraged to explore diverse solutions and share their reasoning without fear of negative judgment. This inclusive atmosphere helped cultivate trust and resilience, especially among students who are less confident in their technical abilities.

Reflections gathered from post-course surveys and focus group interviews (Section 2.4.2) echoed these findings. Many students expressed that the learning environment felt less intimidating and more supportive, and that their confidence in handling technical content had improved. In summary, the observed decline in anxiety scores—particularly those related to terminology difficulty, evaluation pressure, and fear of making mistakes—demonstrates the effectiveness of the IRS–MFL approach in promoting emotional well-being and cognitive preparedness in net-zero carbon education.

The findings from this section offer crucial insights for the practical application of technology in sustainability education.

• Creating a Psychologically Safe Classroom: Educators can use anonymous response systems (like IRS) to create a low-stakes environment for formative assessment. This encourages students to be more open about their knowledge gaps, allowing instructors to address misconceptions immediately. This is particularly valuable in technical and interdisciplinary subjects where students may feel overwhelmed by a large volume of new concepts.
• Optimizing Content Delivery: Course designers should consider a flipped learning model with micro-videos to deliver complex, foundational content. This strategy respects students’ individual learning paces and provides a more effective way to prepare them for hands-on, in-class activities. The reduction in anxiety shows that this approach not only improves comprehension but also contributes to student well-being.
• Integrating Technology for Better Pedagogy: This study demonstrates that technology is not merely a tool for efficiency but a pedagogical asset that can fundamentally change the learning dynamic. The seamless integration of MFL and IRS with VPP simulations creates a coherent and supportive learning journey, from initial concept review to applied problem-solving in a simulated real-world context. This model serves as a blueprint for designing holistic, technology-enhanced sustainability curricula.

3.3. Enhanced Learning Performance Through IRS-MFL Pedagogy

To evaluate the academic impact of the IRS–MFL instructional model, pre- and post-test rubric scores were compared across four key learning dimensions: creativity, creative self-efficacy, learning motivation, and overall learning performance (

Table 3. Pre- and Post-Test Rubric Scores (N = 33).

No.	Dimension	Pre-Test (Mean ± SD)	Post-Test (Mean ± SD)	t-Value
1	Creativity [34]	$2.76\pm 0.48$	$3.39\pm 0.51$	9.32 ***
2	Creative Self-Efficacy [35]	$2.91\pm 0.52$	$3.47\pm 0.49$	8.79 ***
3	Learning Motivation [36]	$2.83\pm 0.61$	$3.45\pm 0.53$	8.41 ***
4	Learning Performance [37]	$3.65\pm 0.51$	$3.85\pm 0.48$	6.12 ***

*** p < 0.001.

). These constructs were directly aligned with the pedagogical goals outlined in Section 2.1 and measured using validated instruments described in Section 2.4.1.

To enhance methodological transparency, the measurement of Learning Performance in this study was further clarified to ensure objectivity and reliability. Rather than relying solely on self-assessment, Learning Performance was evaluated using a composite framework that integrated multiple data sources. Specifically, the final course grade—derived from quizzes, assignments, and project evaluations—served as the primary objective indicator of academic achievement. Additional behavioral evidence, such as attendance and participation records automatically logged through the MFL system, reflected students’ engagement consistency throughout the semester. To complement these objective indicators, students’ cognitive and metacognitive strategy use was measured using a validated subscale of the Motivated Strategies for Learning Questionnaire (MSLQ). This triangulated design combined both objective academic data and process-based indicators, ensuring that the Learning Performance construct represented a robust and balanced measure of students’ achievement and learning processes within the IRS–MFL pedagogical framework.

Among the four categories, creativity demonstrated the most substantial improvement, with average scores rising from 2.76 (SD = 0.48) to 3.39 (SD = 0.51), t = 9.32, p < 0.001. This result suggests that the simulation-based activities and embedded reflective prompts in the IRS–MFL framework effectively stimulated divergent thinking and original idea generation. Prompts such as “What trade-offs were necessary in your dispatch model?” required students to move beyond rote application, encouraging exploration of multiple solutions under carbon constraints. These exercises were further strengthened by peer review mechanisms (see Section 2.2), which prompted iterative improvement through structured critique and collaborative refinement.

Creative self-efficacy also showed notable gains, increasing from 2.91 (SD = 0.52) to 3.47 (SD = 0.49), t = 8.79, p < 0.001. The iterative nature of class simulations, along with visible performance progress through IRS-based feedback, likely contributed to students’ growing belief in their ability to generate effective and novel solutions. The use of anonymous polling helped lower-risk students engage without fear of public error, fostering a psychologically supportive environment conducive to confidence-building.

Learning motivation significantly improved from 2.83 (SD = 0.61) to 3.45 (SD = 0.53), t = 8.41, p < 0.001. This enhancement aligns with instructional observations discussed in Section 2.2, where real-time feedback and gamified IRS quizzes sustained learner engagement. The dynamic adjustment of task complexity—based on live IRS analytics—ensured that students consistently faced appropriately challenging material, maintaining optimal motivational arousal consistent with flow theory principles.

Finally, overall learning performance rose from 3.65 (SD = 0.51) to 3.85 (SD = 0.48), t = 6.12, p < 0.001. Although the effect size was comparatively smaller, this improvement still signifies meaningful gains in applied competence—particularly in carbon dispatch simulation, system modeling aligned with ISO 14064-1:2018, and decision-making under 24/7 CFE constraints. These academic improvements reinforce the effectiveness of the IRS–MFL framework in supporting both conceptual understanding and applied skill development.

Collectively, the statistically significant gains across all four learning dimensions affirm that the IRS–MFL strategy offers a robust instructional approach. By integrating micro-learning content, real-time feedback, peer reflection, and differentiated instruction, the model cultivates not only improved performance but also deeper learner engagement and self-belief in solving complex sustainability challenges.

Taken together, the observed improvements in attendance, reduced anxiety, and academic performance can be interpreted as outcomes of enhanced learning motivation and creative self-efficacy. Motivation fostered consistent participation and preparation, while self-efficacy reinforced students’ confidence in problem-solving and creative application within the VPP simulation context. These findings substantiate the conceptual pathway in which motivation leads to self-efficacy, which in turn enhances performance—an interrelation that underpins the study’s pedagogical framework.

The findings of this study offer several pedagogical insights for designing and implementing sustainability-oriented STEM courses. The integrated IRS–MFL framework demonstrates how a feedback-rich and student-centered learning environment can enhance participation, reduce learning anxiety, and foster creative self-efficacy in complex technical subjects such as carbon management and clean-energy systems.

First, creating psychologically safe learning spaces through anonymous IRS feedback proved effective in encouraging students to participate actively, ask questions, and confront misconceptions without fear of evaluation. This approach is especially beneficial in sustainability education, where interdisciplinary content often intimidates students. By integrating continuous formative assessment, instructors can better monitor affective responses and adapt their teaching dynamically.

Second, blended micro-learning and flipped strategies helped lower cognitive barriers and improved attendance and preparation rates. Short pre-class learning modules enabled students to grasp foundational concepts before class, allowing face-to-face sessions to focus on application and collaborative exploration. This approach can be replicated in other STEM-based sustainability courses that involve abstract or data-intensive topics, such as life-cycle assessment or renewable-energy integration.

Third, collaborative project-based learning anchored in authentic sustainability contexts further strengthened learning motivation and creative confidence. In this study, the Virtual Power Plant (VPP) simulations served as contextualized applications linking the IRS–MFL pedagogy to the Net-Zero Carbon Sustainability objectives. They provided real-world energy-system scenarios in which students could apply theoretical knowledge, practice decision-making, and recognize the interconnectedness of technological, environmental, and policy dimensions. Importantly, the VPP simulations were not treated as independent variables but as pedagogical tools to contextualize learning and make the concept of net-zero more tangible.

Finally, the action-research design adopted in this study highlights the value of iterative instructional refinement within authentic classroom settings. By systematically observing and reflecting on student feedback, instructors can adapt teaching practices to cultivate both engagement and confidence in sustainability-related STEM education. Future applications of the IRS–MFL model may extend to other interdisciplinary domains where cognitive load and anxiety hinder learning—such as circular-economy design, smart-grid engineering, or environmental data analytics.

Overall, the practical implications derived from this study emphasize that technology-enhanced, feedback-driven pedagogies can promote not only academic performance but also the affective and creative capacities essential for achieving net-zero carbon competence. These pedagogical implications collectively bridge the findings with the final conclusions, highlighting how the IRS–MFL framework contributes to both theoretical and practical advancements in sustainability education.

4. Conclusions

This study demonstrates that integrating an Immediate Response System (IRS) with Modified Flipped Learning (MFL) offers an effective pedagogical strategy for enhancing student outcomes in net-zero carbon education. The VPP simulations functioned as thematic applications contextualizing the IRS–MFL pedagogy for sustainability education, rather than as independent variables in the experimental design. The results show that this blended instructional approach improved student attendance, increased active participation, and reduced learning anxiety. Moreover, it significantly enhanced students’ learning motivation, creative self-efficacy, and academic performance—particularly among those with lower confidence in technical subject areas.

By incorporating real-time feedback, peer collaboration, and differentiated pacing, the IRS–MFL framework created a more inclusive and adaptive learning environment compared to conventional instructional models. These findings are consistent with Self-Determination Theory [45], which emphasizes autonomy and competence as key drivers of intrinsic motivation and sustained engagement. In addition, prior research on IRS-based instruction [41] supports the role of interactive feedback in uncovering misconceptions and strengthening conceptual understanding.

In light of the increasing urgency of climate action and the growing demand for carbon literacy in higher education, this study highlights the importance of embedding flexible, student-centered instructional strategies into sustainability curricula. Beyond improving content delivery, the IRS–MFL model effectively addresses emotional and motivational barriers that often hinder engagement with complex and interdisciplinary topics such as carbon management. Future research should further explore the long-term impacts of this pedagogical model on behavioral change and professional readiness, particularly in the context of preparing learners for real-world sustainability challenges.

Future research should further explore the long-term impacts of this pedagogical model on behavioral change and professional readiness. Specifically, we recommend three key directions:

• Application in Diverse Disciplines: Investigate the effectiveness of the IRS–MFL and VPP simulation model in other fields, such as business management, public policy, or the arts, to understand its applicability in a broader context.
• Longitudinal Studies: Conduct follow-up research to assess the long-term retention of knowledge and skills, as well as the sustained impact on student motivation and career choices related to sustainability.
• Mixed-Methods Approach: Integrate qualitative research methods, such as in-depth interviews or ethnographic observation, to gain a richer understanding of students’ lived experiences and perceptions of the learning environment. This would complement the quantitative findings by uncovering the “how” and “why” behind the observed improvements.

Overall, this study contributes to the goals of this Special Issue by providing evidence of measurable learning improvements through an interdisciplinary, technology-enhanced approach. It offers a practical and scalable model for promoting carbon literacy in higher education and advancing the broader mission of sustainability education.