Visual Scaffolding

Artificial Intelligence (AI) has revolutionized educational scaffolding, providing personalized, real-time learning support through machine learning (ML), natural language processing (NLP), reinforcement learning (RL), and computer vision (CV). AI-driven visual scaffolding enhances STEM education, language learning, and special education by offering dynamic feedback and adaptive instruction. However, significant challenges remain, including static scaffolding mechanisms, inadequate calibrated fading, over-reliance on AI assistance, limited metacognitive development, and concerns such as algorithmic bias and data privacy risks. This comprehensive literature review examines the evolution of scaffolding from human-led to AI-driven approaches, evaluates current AI-based implementations, and identifies critical research gaps—particularly in dysgraphia interventions. The study highlights the need for structured fading mechanisms and adaptive feedback to enhance AI-driven scaffolding’s effectiveness and inclusivity. By addressing these limitations, AI-powered scaffolding can transition from rigid, rule-based interventions to truly adaptive learning support systems, fostering cognitive development, independent problem-solving, and long-term knowledge retention. Future research should focus on integrating AI-driven adaptive scaffolding, implementing structured fading strategies, and conducting longitudinal studies to assess AI’s sustained impact on learning outcomes.

No abstract available

No abstract available

The main purpose of this research is to analyse the students' respond toward Visual Scaffolding activity in the EFL classroom. The data in this study is taken from twenty five EFL students’ reflection they have narrated based on their class participation about Visual Scaffolding activty. The researcher collected the results of their reflections through Microsoft Word documents that have been shared by the researcher before they wrote their reflective journal. Then, those reflective journal were analyzed by using thematic analysis (TA) adopted from Braun and Clarke (2006) through coding the data by underlining some words as a code. The results of the research revealed that the students perceive visual scaffolding positively and enthusiastically since it was able to support them to understand the presented material throughout the teaching and learning process in the class. The students also felt that visual scaffolding could  motivate them to study. On the other hand, there were a few students stated that they were not feel any positive changes after participating Visual Scaffolding activity in their class but they admit that Visual Scaffolding can help them to understand the learning material properly.

One approach to training is using scaffolding to assist the novice trainee in performing tasks while they develop fundamental skills. This paper describes the implementation and testing of visual scaffolding to support training of attentional management skills. The training tool was developed specifically for sensor operators of unmanned aerial systems (UAS) to learn attentional skills associated with managing the competing demands of UAS operational tasks. A key part of the training included attentional scaffolding—visual highlighting that was used to guide the trainee’s attention. Three different evaluations were conducted to assess performance while leveraging scaffolding: Two were performed with U.S. Navy and U.S. Coast Guard (USCG) sensor operators, and one was performed in a human-in-the-loop (HITL) study. In this paper, we detail the Navy and USCG studies, summarize the HITL study, and synthesize findings across the evaluations.

Visual scaffolding, a type of scaffolding provided in a visual format, is considered to have potential in web-based learning environments as it naturally includes tacit information, highlights the critical features of tasks and promotes learning through the spatial organization of tasks. However, systematic approaches to visual scaffolding design principles are rare. This study aims to develop visual scaffolding design principles and guidelines in a web-based learning environment through a literature review. The specific visual scaffolding design principles regarding objects are 1) the spatial organization of information, 2) coaching, modelling, and mentoring, 3) explication of the task structure, 4) explication of the process, 5) explication of the cognitive process, 6) explication of understanding, 7) metacognitive prompts. In terms of visuals, the principles drawn are 8) ensuring visibility, coupling, and usability, 9) the use of appropriate visuals, 10) proximity and simplicity of the display, 11) spatial organization of the toolset. The principles regarding the learners are 12) consideration of different learners, and 13) manipulation of support and use. The limitations and future research directions are discussed.

Visual problem-solving is an essential skill for professionals in various visual domains. Novices in these domains acquire such skills through interactions with experts (e.g., apprenticeships). Experts guide novice visual problem-solving with scaffolding behaviours. However, there is little consensus about the description and classification of scaffolding behaviours in practice, and to our knowledge, no framework connects scaffolding to underlying cognitive mechanisms. Understanding effective scaffolding is particularly relevant to domain-specific expert-novice research regarding visual problem-solving, where in-person scaffolding by an expert is a primary teaching method. Scaffolding regulates the flow of information within the learner’s working memory, thereby reducing cognitive load. By examining scaffolding research from the perspective of cognitive load theory, we aspire to classify scaffolding behaviours as cognitive behaviours of cueing (which involves attention allocation) and chunking (the practice of grouping information, often in conjunction with prior knowledge), into a cohesive and unified framework. In this scoping review, 6533 articles were considered, from which 18 were included. From these 18 articles, 164 excerpts describing expert-novice interaction were examined and categorised based on cognitive strategy (cueing or chunking) and method of expression (verbal or nonverbal). An inductive category (active or passive) was also identified and coded. Most scaffolding behaviours were categorised as active verbal cueing and active verbal chunking. Qualitative patterns in excerpts were collated into 12 findings. Our framework may help to integrate existing and new scaffolding research, form the basis for future expert-novice interaction research, and provide insights into the fine-grained processes that comprise scaffolded visual problem-solving.

Penelitian ini menganalisis potensi pengembangan Visual Representational Scaffolding Model (VRSM) sebagai inovasi pembelajaran bernilai ekonomi dalam konteks edupreneurship dan penguatan peran guru sebagai teacherpreneur. Pendekatan yang digunakan adalah konseptual deskriptif melalui kajian literatur tahun 2023 sampai 2025. Hasil kajian menunjukkan bahwa VRSM tidak hanya efektif meningkatkan pemahaman konseptual dan pemecahan masalah siswa, tetapi juga dapat dikembangkan menjadi produk digital edukatif yang memiliki nilai komersial. Integrasi teknologi seperti GeoGebra dan AR/VR memperkuat posisi VRSM dalam ekosistem edutech berkelanjutan. Kajian ini menghasilkan sintesis konseptual dan model BMC sederhana yang menegaskan hubungan antara inovasi pedagogis, nilai ekonomi pendidikan, dan praktik teacherpreneurship yang mendukung transformasi digital pendidikan abad ke-21.

This research is motivated by the fundamental obstacles faced by mild intellectually disabled students in mastering prayer practices due to deficits in working memory and sequential thinking skills. The main problems include dependence on verbal instructions and low memory retention of movement sequences. This study aims to analyze the implementation of visual scaffolding media as a cognitive scaffold at SLB Mutiara Bunda Cirebon. The research method used is descriptive qualitative. The results showed that the use of permanent visual cues was able to reduce students' cognitive load and minimize prompt dependency. Through the fading strategy, students showed an increase in independence and order of movement sequences (tumakninah). The integration of this media is proven effective in bridging the transition from full guidance to worship independence

The objectives of  this study are to find out whether or not;  1) there is a significant increase in students’ reading comprehension achievement based on after they are taught by using visual scaffolding strategy and intensive reading strategy and those who are not. 2) there is a significant contribution of visual scaffolding strategy and intensive reading strategy towards reading comprehension achievement  of the high, average, and low achievers? This reserach used factorial experimental designs.  These designs are essentially modifications of either the posttest-only control group or pretest-posttest control group designs. The sample was the students of PGRI University of Palembang. Two classes were chosen in which first class was experimental with male students was applied and the second was the second experiemnetal group trained by intensive reading strategy with the female students. The results showed that First, Visual scaffolding and Intensive Learning Strategy technique affected the students’ reading achievement of PGRI University. The result of t independent test was the mean of experiment group higher than mean of control group. It could be concluded, experiment group and control group were significantly difference and the visual scaffolding and Intensive Learning Strategy technique was effectively to use in improving students’ reading achievement.  The Second, there was an interaction effect of visual scaffolding and intensive learning strategy technique and students’ learning achievement toward the students’ reading achievement.

As programming education gains global importance, learners often struggle due to the cognitive demands, particularly in textual environments. Visual programming tools have been developed to ease this burden, but learners’ cognitive abilities—specifically executive function (EF)—may still affect their programming process and strategies. This study investigates how visual scaffolding influences the programming strategies of learners with different EF levels. Thirty-two adult participants were randomly divided into experimental and control groups, using programming systems with and without visual scaffolding respectively. Their EF was assessed using the Simon and running span tasks. Also, an eye-tracker was used to explore their cognitive processes during programming tasks and to identify their programming strategies. Results show that high-EF learners tended to adopt top-down strategies regardless of scaffolding. In contrast, low-EF learners without scaffolding used mixed strategies, while those with visual scaffolding were guided toward adopting a top-down approach. These findings highlight the role of visual scaffolding in supporting low-EF learners by fostering more effective programming strategies and mitigating cognitive differences between high- and low-EF learners.

Penelitian ini bertujuan untuk mengetahui pengaruh model pembelajaran explicit instruction berbasis visual scaffolding terhadap kemampuan numerasi siswa UPT SPF SD Negeri 104247 Tanjung Mulia. Jenis penelitian ini adalah penelitian eksperimen dengan pendekatan kuantitatif quasi experimental tipe none equivalent control group desain. Subjek penelitian ini adalah seluruh siswa kelas 2 SD Negeri 104247 Tanjung Mulia yang berjumlah 58 siswa dari 2 rombongan belajar.Hasil penelitian menunjukkan bahwa rata-rata skor pretest siswa adalah 51,5, kemudian setelah menerapkan model pembelajaran explicit instruction berbasis visual scaffolding diperoleh rata-rata skor posttest siswa sebesar 82,59. Selanjutnya hasil uji hipotesis menggunakan paired sample t-test diperoleh nilai sig.(2-tailed) = 0,000 < 0,05. Dengan demikian H₁ diterima H₀ ditolak. Hal tersebut menunjukkan bahwa rata-rata hasil belajar siswa setelah penerapan model pembelajaran explicit instruction berbasis visual scaffolding berpengaruh terhadap kemampuan numerasi siswa UPT SPF SD Negeri 104247 Tanjung Mulia.

This study aims to verify the effects of scaffolding on learning outcomes in the acquisition of visual-to-auditory substitution skills and to analyze learners' perceptions of four types of scaffolding. To achieve this, an instructional program was designed based on Cognitive Load Theory and Scaffolding Theory to manage high cognitive load and facilitate learning. An experimental study was conducted with 45 adult learners, employing an explanatory sequential mixed-methods design. To analyze the differences in learning outcomes between the control group and experimental group, the provision of scaffolding was set as the independent variable and learning outcomes as the dependent variable. An Analysis of Covariance(ANCOVA) was performed, controlling for prior musical ability. The results revealed that while no significant difference existed between the groups at pre-test, the experimental group demonstrated statistically significant gains in post-test learning outcomes. Furthermore, a repeated measures ANCOVA confirmed a significant interaction effect over time. A qualitative analysis of the four types of scaffolding indicated that all types were perceived positively, with metacognitive scaffolding reported as the most effective. These findings suggest that scaffolding is an effective instructional strategy for managing the high cognitive load associated with novel and non-intuitive tasks, thereby promoting learners' self-regulated learning.

Large language models (LLMs) are increasingly used by students, yet many novice users struggle to construct effective prompts because prompt engineering is rarely taught explicitly. As a result, novices often lack guidance on translating their intentions into effective interactions with LLMs. Prior work frames prompt-based interaction as a gulf of envisioning, where users struggle to translate abstract goals into concrete prompt specifications, express intentions in model-interpretable ways, and assess generated outputs. To address this gap, this paper introduces PromptFlow, a node-based prompt-scaffolding system that externalizes prompt structure into modular components and supports explicit specification and comparison of prompt variants. A user study with novice participants examined how such interface-level scaffolding influences prompt-engineering practices. Results suggest increased awareness of prompt components, evidence of strategy transfer when participants returned to a standard chat-based interface, and perceived usefulness of scaffolding features for addressing challenges in conventional chat-based prompting.

No abstract available

No abstract available

Arabic writing skill (maharah al-kitabah) remains a challenge for madrasah students due to limited vocabulary mastery and difficulties in organizing ideas into coherent texts. This study aims to examine the effect of the Picture and Picture strategy on improving Arabic writing skills among ninth-grade students at MTs Nurul Iman Ulu Gedong, Jambi. This research employed a quantitative approach using a quasi-experimental method with a Non-Equivalent Control Group Design involving 30 students divided into experimental and control groups. Data were collected through writing tests administered as pre-test and post-test and analyzed using paired samples t-test and independent samples t-test. The results showed that both groups experienced significant improvement; however, the experimental group achieved higher improvement with a mean difference of 23.27 (t(14) = -14.530, p < 0.001) compared to the control group with a mean difference of 10.57 (t(14) = -6.764, p < 0.001). The independent samples t-test confirmed a significant difference between the two groups (t(28) = 7.586, p < 0.001). The novelty of this study lies in revealing that Picture and Picture functions as visual scaffolding that supports students in organizing ideas, developing vocabulary, and constructing Arabic sentences systematically. This study theoretically strengthens visual-based approaches in Arabic language learning and practically provides an alternative strategy for teachers to improve students’ writing competence.

I describe a kinesthetic activity about oxygen handling by hemoglobin with two specific goals: 1) to help students gain a better understanding of how hemoglobin properties affect oxygen delivery and 2) to improve the ability of the students to actually read the hemoglobin oxygen-binding curve. The activity makes understanding oxygen delivery more intuitive, provides a kinesthetic analog to delivery of oxygen, and provides data to plot for the hemoglobin-oxygen curve.

The present study aims at:1- Finding out the effectiveness of using visual scaffolding strategy on students‟ achievement in English.2- Finding out whether there is any statistically significant difference between experimental group achievement at the recognition level and at the production level.3- Finding out whether there is any significant difference between experimental group's achievement in the pre-test and post-test.To verify the hypotheses of the study, a sample of (70) students have been randomly selected from first intermediate class at Al Nahreen intermediate school in Al-shirqat city, and divided into two groups, i .e the experimental (35) and the control group(35). Both groups are equalized in their age and their parents‟ level of education. An achievement test of five questions has been constructed, validated, its reliability is obtained, its data analyzed and then applied to each groups of study. The data has been statistically analyzed and show the following results:1. There are statistically significant differences between the achievement of the two groups and in favour of the experimental group.2. The statistically significant differences between the student‟s achievement at the production level and their performance at the recognition level are equal.3. There are statistically significant differences between the students‟ performance in the experimental in the pretest and their performance in the post test in favour of visual scaffolding.

Fostering metacognitive awareness and reflective thinking remains a challenge in English as a Foreign Language (EFL) writing instruction. These cognitive processes are essential for enabling learners to monitor, evaluate, and engage in meaningful self-reflection throughout the writing cycle. Using Flower and Hayes’s cognitive process model and Jasper’s Experience-Reflection-Action framework, this study proposes a Generative AI-assisted Visually Guided Progressive (GAI-VGP) writing approach. This pedagogical design integrates step-by-step visual scaffolds with interactive AI feedback to foreground learner agency and promote reflective engagement. To evaluate this approach, a semester-long quasi-experimental study was conducted with 58 EFL university students. Data sources included pre- and post-writing tests, metacognitive awareness questionnaires, reflective thinking scales, AI chat logs, and reflection notes. The results showed that the GAI-VGP group significantly outperformed the conventional VGP group in writing performance, planning and evaluation strategies, and reflective awareness. Qualitative analyses indicated increased engagement in understanding and reflectively evaluating writing strategies, although critical reflection remained limited across both groups. Although AI significantly enhanced planning and evaluation, monitoring showed no significant gains, with prompt-log evidence revealing a displacement pattern in which learners outsourced real-time regulation to AI validation.

No abstract available

Introduction While engaging in knowledge acquisition in online learning environments, especially online text reading, students tend to be disoriented and face a cognitive load problem (Chen & Dwyer, 2003). Under this situation, specific instructional scaffolds should be used to assist learners to comprehend large amount of information (Sharma & Hannafin, 2007). Of those available scaffolds listed in the existing literature, visual scaffolds are regarded as an effective tool to support student learning because human beings are visually oriented (Dwyer, 2007; Norman, 2004). According to Davis (2007), the text-based reading document is still a mainstream instructional material in online learning settings despite the availability of high technologies. As Chen and Dwyer (2003) pointed out, online instructors often employed hypermedia documents as reading materials to support student learning. In order to pursue better instructional effectiveness, high-quality design in online reading materials is necessary. However, whether or not embedding visual scaffolds can enhance students' online reading performance is worthy of exploration. For the background information discussed above, this study aims to explore the in-instructional effectiveness of different types of visual scaffolds embedded in online reading material. Engineering students were chosen as a targeted group. One control group with no visual scaffolds was created. Two visuals, static and interactive visual scaffolds, were developed in an experimental study and serve as two treatment groups. The online reading material was a website that imparts basic science knowledge about the human heart. A post-test, consisting of identification, terminology, and comprehensive tests, was used to measure students' online reading performance. Specifically, the purpose of the study is to evaluate the instructional effectiveness of two types of visual scaffolds for engineering students in an online reading environment. Theoretical Foundation The Concept of Scaffolding Scaffolding is defined as "an adult controlling those elements of the task that are essentially beyond the learner' capacity, thus permitting him to concentrate upon and complete only those elements that are within his range of competence" (Wood, Bruner, & Ross, 1976, p.9). A theoretical foundation behind scaffolding is Vygotsky's zone of proximal development, which emphasizes the role of social interaction in promoting cognitive development and bridging the gap between what learners actually know and potentially know (Sharma & Hannafin, 2007). To date, scaffolding no longer restricts interactions between individuals. It has extended to technological tools and instructional techniques, which are often regarded as scaffolds (Puntambekar & Hubscher, 2005). For example, Zumbach, Reimam, and Koch (2006) designed a feedback-based instructional technique to promote a higher-level thinking during online discussions. The Feature of Scaffold In online learning settings, Hill and Hannafin (2001) proposed four types of scaffolds--procedural, strategic, metacognitive, and conceptual scaffolds--which can be employed to support student learning. Procedural scaffolds assist learners to use online resources and to "clarify requirement and reduce cognitive load" (p. 45). Strategic scaffolds provide learners with alternative approaches to engage online tasks. Metacognitive scaffolds allow learners to "assess what they know and what to do as they learn" (p. 45). Conceptual scaffolds help learners digest online information and facilitate knowledge construction. In directed distance learning environments (DDLEs), Sharma, Oliver, and Hannafin (2007) defined DDLEs scaffolds as "planned strategies and content structures that assist the learner in more efficiently and effectively processing and internalizing course materials" (p. 265). Basically, DDLEs scaffolds are instructional techniques that aim to improve students' learning performances in terms of knowledge acquisition. …

Generative AI (GenAI) poses a pedagogical challenge to data visualization education by enabling the rapid creation of persuasive, misleading visuals. This design case presents a three-phase learning experience designed for a graduate course to foster critical AI-visual literacy. The design inverts the typical use of AI by tasking students with using GenAI as a ‘cognitive partner’ to intentionally create a misleading version of their own honest data visualization. The activity guides students through establishing an ethical baseline, creating the misleading visual, and engaging in peer critique, supported by scaffolds like a prompting guide and reflection templates. Analysis of student work shows the design effectively fostered critical AI literacy.

Students in Higher Education (HE) with Attention Deficit Hyperactivity Disorder (ADHD) often face persistent challenges in video-based e-learning, including difficulties with sustained attention, comprehension, and information retention. Existing lecture platforms typically offer limited embedded support, while external tools for focus and note-taking operate outside the learning context and rely heavily on self-regulation. This paper presents the design, development, and evaluation of a digital learning tool that integrates support features directly into the lecture interface to assist students with ADHD. Developed using a user-centered design methodology, the prototype includes chapter-based video segmentation, AI-generated summaries and quizzes, a context-aware chatbot, and visual scaffolds such as mind maps and annotated slides. The design was informed by the input of students with ADHD and experts in Education, Accessibility, and Human-Computer Interaction (HCI). A high-fidelity prototype has been developed and evaluated to assess usability, relevance, and its potential to support focus, comprehension, and retention. The findings suggest that embedding accessible adaptive features in lecture content can reduce cognitive load and improve engagement for students with ADHD. The paper concludes with implications for inclusive e-learning design and directions for future research.

No abstract available

This mixed-methods study examines whether backpropagation-based deep learning (DL) visualizations can strengthen metacognition and learning outcomes in a university Linear Programming course. Sixty undergraduates (8-week blended format) completed pre/post cognitive tests and the Metacognitive Awareness Inventory (MAI), while their LMS activity traces (e.g., time-on-task, revision frequency, error types) trained a multilayer perceptron. The intervention exposed students to DL visual artifacts—loss curves, gradient/weight updates, and error heatmaps—as reflective scaffolds linking machine error correction to human self-regulation. Quantitatively, mean test scores increased from 61.23 to 80.57 (paired-t, p < .001), and total MAI rose from 135.40 to 159.85 (paired-t, p < .001). Gains concentrated in regulation of cognition (monitoring/evaluation). Metacognitive improvement correlated with achievement (Pearson r = .62, p < .001). Computationally, model loss decreased from 0.25 to 0.03 over 200 epochs with 89.4% validation accuracy; Dynamic Time Warping = 0.81 (p < .01) indicated strong temporal alignment between DL loss minimization and students’ learning curves. Qualitatively, thematic analysis of weekly reflections and interviews revealed a progression from error recognition to strategy adjustment and reflective transformation, recasting errors as actionable signals. Triangulating quantitative, computational, and qualitative strands supports the Cognitive Backpropagation Learning (CBL) framework: DL error feedback parallels human metacognitive feedback, and its visualization functions as a digital mirror that externalizes reflection. Findings recommend interpretable DL dashboards as practical, class-deployable scaffolds to cultivate reflective, adaptive mathematical thinkers.

This paper presents the design and implementation of a digital self-paced learning format that scaffolds both knowledge acquisition and time management. While self-paced environments offer flexibility and learner autonomy, they also pose challenges in sustaining motivation, managing cognitive load, and regulating study behaviour. The learning environment supports learners through two key scaffolding layers: content-level guidance and temporal regulation.Knowledge scaffolding is achieved through structured content design, semantic navigation tools, and Retrieval-Augmented Generation (RAG) that provides personalized summaries based on learner performance. These features help learners build conceptual understanding and reinforce key concepts. Temporal scaffolding includes time-aware notifications, visual progress dashboards, and learning caps to encourage regular engagement and reduce last-minute cramming.Empirical data from over 600 learners demonstrate improved study behaviour and performance with structured guidance. While artificial intelligence plays a vital role in personalization and feedback, it must maintain transparency and trust. The system is designed to act as a supportive companion, not a controlling presence—preserving learner autonomy and ownership.This work highlights the potential of AI-augmented scaffolding to create human-centred, effective self-paced learning environments.

Students with disabilities, including those with learning disabilities (LD), face persistent challenges in acquiring and retaining academic vocabulary, which can undermine their academic success across content areas and overall reading achievement. Content Acquisition Podcasts (CAPs) are multimedia instructional tools that deliver explicit, evidence-based vocabulary instruction in accessible, engaging formats. CAPs are a flexible tool that can be used to support vocabulary acquisition and comprehension, especially in inclusive settings where students with LD are receiving instruction and would benefit from structured supports, visual scaffolds, and repeated exposure. This article reintroduces CAPs, outlines recent enhancements to their design, provides step-by-step procedures for creating CAPs, and offers actionable suggestions for implementing CAPs in science, social studies, English, and mathematics classrooms. Overall, this article outlines how CAPs provide a flexible approach to supporting students with LD through research-based multimedia instruction that fosters vocabulary acquisition and deeper conceptual understanding across various academic content areas.

No abstract available

This study investigates the effectiveness of applying Sweller’s (1988) cognitive load theory (CLT) principles to improve the learning of derivative concepts among first-year university students. The research employed a quasi-experimental pre-/post-test control group design to evaluate the impact of CLT-based instructional strategies–worked examples, visual scaffolds, and structured problem-solving–on students’ achievement and perceived cognitive load. A total of 60 students were randomly assigned to either an experimental group, receiving CLT-based instruction, or a control group, receiving traditional teaching. Data were collected using achievement tests and the cognitive load scale, assessing intrinsic, extraneous, and germane load components. Quantitative data were analyzed using independent samples t-tests and two-way and mixed-design ANOVA, with effect sizes (Cohen’s d, partial η²) reported to determine the magnitude of observed effects. Results indicated that the CLT-based instruction led to significantly higher achievement scores, reduced extraneous cognitive load, and increased germane processing compared to traditional methods. The findings support the theoretical validity of CLT and its practical applicability for mathematics instruction, suggesting that structured visual and example-based learning can optimize students’ cognitive resources and promote deeper understanding of calculus concepts.

This study investigates the types of scaffolding strategies employed by elementary teachers during read aloud sessions, with a specific focus on emergent bilingual students. Through a multi‐site explanatory case study involving 14 elementary teachers across four districts, we analyzed the frequency and quality of scaffold use, categorizing them into interactive, linguistic, visual, and kinesthetic types. Findings reveal that interactive and linguistic scaffolds were the most frequently used (i.e., n = 446 for interactive, n = 399 for linguistic), often characterized by high teacher control (e.g., question–answer patterns), while visual and kinesthetic scaffolds (i.e., n = 155 for visual, n = 126 for kinesthetic) were underutilized. The study highlights the importance of coherent and congruent scaffold use, emphasizing that the quality of scaffolding, rather than the quantity, shapes student engagement and comprehension. The results suggest that read alouds, when strategically scaffolded, can serve as a powerful tool for enhancing both language and content learning. Implications highlight the need for teacher training in dynamic, student‐responsive scaffolding practices.

Artificial intelligence (AI) is reshaping education by enabling adaptive and personalized experiences. This study explores the integration of Gemini AI’s text-to-image generation into an adaptive learning environment for undergraduate BSIT students. Using a mixed-method developmental design with 100 participants, we analyzed learning gains through pre/post testing, motivation via the IMMS survey, interaction telemetry, and focus group discussions (FGDs). Quantitative analyses revealed substantial improvement from pre-test to post-test, with a significant paired t-test and an ANOVA indicating differences across performance subgroups. Post-hoc comparisons (Tukey HSD) suggested that lower-performing learners benefited most, pointing to the potential of visual scaffolds to narrow achievement gaps. Qualitative findings supported these results, highlighting enhanced conceptual understanding, stronger motivation through co-creation, and increased teacher trust when explainability accompanied AI-generated visuals. Benchmarking showed a trade-off: Gemini provided higher-fidelity visuals suited to complex concepts, while a lightweight variant (NanoBanana) offered lower latency and broader device accessibility. A hybrid routing approach is conceptually recommended to balance fidelity and speed; however, hybrid evaluation was proposed but not executed and is reserved for future empirical testing. Overall, the findings support text-to-image generation as a practical means to strengthen adaptive learning and advance inclusive, equitable quality education aligned with SDG4.

This study addresses the pedagogical challenges in multilingual elementary classrooms by designing and validating an illustrated vocabulary learning resource that integrates local wisdom for students in Flores, Indonesia. The research was motivated by the lack of structured materials supporting the simultaneous acquisition of local, Indonesian, and English vocabulary, particularly in contexts where visual and contextual scaffolds are limited. We employed a Research and Development (R&D) approach guided by the ADDIE model, collecting data from six elementary schools across three linguistic communities (Lio, Sika, Ngada) through classroom observations, teacher interviews, and student questionnaires. The developed resource, created using Canva, features ten thematic units with vocabulary presented in three languages alongside culturally relevant illustrations. Expert validation confirmed high content validity (Lio: 1.00; Ngada/Sika: 0.86) and mean scores above 4.60 on a 5-point scale, categorizing the resource as “very good.” User evaluations further demonstrated strong perceived usefulness, with mean scores ranging from 4.80 to 4.93 for clarity, visual presentation, and multilingual structure. The results indicate that the integration of visual dual-channel processing, multilingual scaffolding, and culturally contextualized content significantly enhances vocabulary acquisition and engagement. This study contributes a systematically validated instructional framework that bridges cognitive and contextual learning principles, offering a practical solution for multilingual education in linguistically diverse settings. The findings underscore the importance of culturally responsive materials in fostering equitable learning opportunities for elementary students.

Social robots are receiving an ever-increasing interest in popular media and scientific literature. Yet, empirical evaluation of the educational use of social robots remains limited. In the current paper, we focus on how different scaffolds (co-speech hand gestures vs. visual cues presented on the screen) influence the effectiveness of a robot second language (L2) tutor. In two studies, Turkish-speaking 5-year-olds (n = 72) learned English measurement terms (e.g., big, wide) either from a robot or a human tutor. We asked whether (1) the robot tutor can be as effective as the human tutor when they follow the same protocol, (2) the scaffolds differ in how they support L2 vocabulary learning, and (3) the types of hand gestures affect the effectiveness of teaching. In all conditions, children learned new L2 words equally successfully from the robot tutor and the human tutor. However, the tutors were more effective when teaching was supported by the on-screen cues that directed children's attention to the referents of target words, compared to when the tutor performed co-speech hand gestures representing the target words (i.e., iconic gestures) or pointing at the referents (i.e., deictic gestures). The types of gestures did not significantly influence learning. These findings support the potential of social robots as a supplementary tool to help young children learn language but suggest that the specifics of implementation need to be carefully considered to maximize learning gains. Broader theoretical and practical issues regarding the use of educational robots are also discussed.

Graphic organizers (GOs) such as Venn diagrams are tools that help instructors support students at all levels of learning in making sense of content. What happens when the readily available GOs or similar tools do not meet the needs of your specific class or your specific content? When we couldn’t find a tool that scaffolded our students’ ability to create language objectives, we created our own content and context-specific graphic organizer tool as a way to mediate student understanding of a complex concept, specifically a form/function analysis of clauses. This paper describes students’ need for scaffolds and our process of creating a very specific tool for analyzing clauses and sentences. We also lay out the steps others can take to create their own tool for their classroom. These steps include guiding principles and questions to support faculty across higher education disciplines and contexts in their own analysis and creation of scaffolds for their classrooms.

No abstract available

This study examined the effectiveness of digital storybooks and graphic organizers integrated with the ESP framework in enhancing narrative reading comprehension among vocational high school students in Digital Business field. Employing a quasi-experimental design, the study included 61 eleventh-grade students divided into experimental and control groups. Data were obtained from pre-test and post-test scores and analyzed through paired-sample and independent-sample t-tests. The experimental group, which received treatment, presented a significant mean score increase from 46.13 to 70.77, while the control group slightly improved from 41.50 to 50.63. These results indicate that the integration of digital storybooks and graphic organizers could effectively enhance students’ reading comprehension. Grounded in Multimedia Learning Theory and visual scaffolding, combining interactive digital tools with structured visual aids is an effective strategy for teaching narrative texts which support learners’ comprehension through multimodal input and structured visual representation. This research contributes to ESP pedagogy by demonstrating the integration of digital storybooks and graphic organizers within vocational Digital Business contexts supports narrative reading comprehension tailored to students’ occupational needs. Future studies could further explore the long-term impact of such tools, their application across different text genres, and their potential for enhancing other language skills.

No abstract available

No abstract available

No abstract available

This study examines the relationship between instructional scaffolding and metacognitive strategies of language teachers in selected public secondary schools in the Division of Island Garden City of Samal. A descriptive-correlational research design was used with a sample of 132 language teachers. Data were collected through standardized questionnaires administered via face-to-face surveys. The data were analyzed using mean, standard deviation (SD), Pearson product-moment correlation, and multiple linear regression analyses. The results revealed that both instructional scaffolding and metacognitive strategies were rated very high. Correlation analysis indicated a significant positive relationship between these variables. Furthermore, specific domains of instructional scaffolding, including situation, visual, question stems, graphic organizers, and processing, were found to significantly influence teachers' metacognitive strategies. It is recommended that school administrators focus on further strengthening instructional scaffolding to enhance the development of metacognitive strategies among teachers. Promoting teacher collaboration, providing professional development opportunities, and encouraging reflective practices will help sustain strong levels of metacognitive strategies and improve teaching effectiveness in schools.

ABSTRACT Situated within the writer(s)-withincommunity (WWC) model of writing and the Recursive Writing Model, this study examined the planned scaffolds of nine teachers during writing instruction with 3rd- through 5th-grade students. Data collection included four approximately 60-minute observations of classroom instruction per teacher and an interview with each teacher about their approaches to writing instruction. Qualitative analysis revealed that teachers used a variety of scaffolds to support students’ writing and that teachers’ use of various scaffolds was contingent on whether they were supporting students during the We Study or the We Write stage of the Recursive Writing Model. Specifically, during the We Study stage, teachers scaffolded students through the use of mentor texts and discussions that highlighted the functions and features of these texts. During the We Write stage, teachers selected and sequenced tasks to provide incremental support, provided students with tools such as graphic organizers to facilitate their writing, and used various participant structures, including one-on-one check-ins, to monitor the writing process and provide individualized contingent feedback. Findings contribute to research on how teachers scaffold and gradually release responsibility during writing instruction.

This research focuses on how designing, implementing, and evaluating didactic strategies and activities based on scaffolding with visualization and the use of graphic organizers guide the process of reading comprehension in an eighth-grade public school course placed at A1 level. The research methodology was framed as action research with a mixed-method approach and a quasi-experimental design with no control group. There were 31 participants aged 12 to 18. The research design implied (1) a pretest, (2) a diagnostic analysis, (3) intervention, (4) a posttest, and (5) final analysis. After gathering information from the pretest initial diagnostic data, the pedagogical intervention with ten task-like scaffolding activities was applied and assessed with rubrics. Posttest data were collected and accounted for from the researcher’s journal notes. Both reading comprehension tests, the pretest, and the posttest had 21 questions divided into literal, inferential, and evaluative levels. Results obtained using descriptive statistics exposed, from the posttest, some improvement in the literal and evaluative level questions but a slight decline in inferential level questions contrasted with the pretest results. Data demonstrated that some external factors like time, academic aids, and responsibility influenced the outcome positively or negatively.

This article describes a comprehensive framework (iDISC) for the close reading of informational texts for elementary students that may need additional language, social, memory, or behavioral supports. The article introduces concrete tools that are used before, during, and after close reading, including cue‐cards, language stems, discussion behaviors, anchor posters, and graphic organizers. The tools provide teachers with instructional scaffolds that can support their students to undertake more educative discussions with peers.

Against the backdrop of the current English curriculum's integration of instrumentality and humanity, this paper explores the application of scaffolding instruction to senior high school English reading instruction. It first outlines the core concept and theoretical foundations of scaffolding instruction, then elaborates on the concrete implementation procedures, and analyzes its pedagogical advantages in real classroom practice. Finally, supported by specific teaching cases, the paper focuses on the practical use of question scaffolds and graphic organizers in senior high school English reading instruction.

This study aims to self-reflect the student-teacher’s (ST’s) scaffolding during her online teaching practicum practice. The self-reflective was conducted in six meetings using a Zoom Application. It utilized a thematic analysis from Braun Clarke (2006). The findings showed that the student-teacher was able to: (1) be aware of her scaffolding drawbacks and improved significantly in the next meetings, (2) improve students’ understanding, (3) conduct appropriate scaffolding learning process by utilizing various graphic organizers and media. While the student teacher was effective in using the scaffolding technique, each theme's improvement path is different. Every meeting is typically used to develop it, and it is acquired in a step-by-step manner.

Developing effective writing proficiency continues to pose significant difficulties for EFL learners, often stemming from limited engagement, inadequate scaffolding, and traditional teacher-centered approaches. Thus, this present mixed-methods study explores how students perceive the restructured Interactive Dynamic Literacy Model (IDLM)-driven Problem-Based Learning (PBL) framework, scaffolded by graphic organizers, to cultivate interactive literacy and advance students’ writing development. Using a mixed-methods approach, data were systematically gathered over a four-week implementation involving 25 university EFL students through a validated Likert-scale questionnaire and semi-structured interviews. The quantitative findings showed generally favorable perceptions of the instructional model (overall M = 3.36), especially regarding the learning syntax, reaction principles, and the instructional impact. The scale’s reliability coefficient (Cronbach’s α = 0.625) falls within an acceptable range for instruments used in exploratory measures capturing multiple perception constructs. The thematic analysis of interview data further illuminated how the redesigned instructional model facilitated learning, supported students' cognitive engagement, idea generation, collaborative inquiry, and constructive use of feedback. Even so, the analysis also revealed persistent challenges, including group dynamics, inconsistent use of assessment tools, and limitations in the instructor’s ability to provide equitable real-time guidance. Integrating both data strands, the study demonstrates that the IDLM-based PBL framework with graphic organizer support fosters a supportive, engaging, and socially interactive environment that promotes writing development, while also revealing the need for differentiated scaffolding and adaptive instructional practices to address diverse learner needs. Implications for writing pedagogy, curriculum development, and avenues for further inquiry.

Introduction. Writing skill is one of the most important skills that English-majored students need to master to apply to their jobs after graduation. However, developing writing skill is not easy, especially for students who are not native speakers. To solve that problem, instructional scaffolding plays a very important role to help English-majored students improve their writing skills. Instructional scaffolding is an approach to teaching that involves providing students with temporary support and guidance as they learn new concepts and skills. The goal of scaffolding is to gradually remove this support as students become more proficient, so they can eventually work independently. This approach can be used in a variety of educational settings, from early childhood education to higher education. Scaffolding can take many forms, such as visual aids, graphic organizers, guided questions, or peer collaboration. The type of scaffolding used depends on the learning objectives and the needs of the individual student. In short, instructional scaffolding is a powerful tool for promoting student learning and success. By providing targeted support and guidance, educators can help students build their knowledge and skills in a structured and effective way. In this article, the author will focus on researching the role of scaffolding in writing skills for English majored students at The University of Khanh Hoa. Study participants and methods. The article uses the method of questionnaire survey and in-depth interview to study the status of the role of instructional scaffolding in writing skills of English-majored students at The University of Khanh Hoa. Besides the author also uses methods such as: analysis; synthetic; logic; history to write articles. To have an objective view of the role of academic support activities in writing skill of English-majored students at the University of Khanh Hoa, the author conducted a survey of 133 students studying “writing 1”, in the academic year 2022-2023 and selected 13 students to conduct in-depth interviews. The design is done on a five-step scale. The survey results were processed by SPSS software, with the question why instructional scaffolding are necessary in the English writing classes? The results obtained provide vocabulary and grammar that students agree with at a high level, reaching a level above 4.5, showing that learners appreciate scaffolding. Activities such as doing exercises are 3.47; students' cross feedback is 3.87; the lecturer provides sample text with the corresponding mean value of 3.8; Students reading specialized documents and articles is 3.76. From the data obtained, it is shown that the provision of vocabulary and grammar is the activity most appreciated by students in scaffolding. Meanwhile, doing the set of exercises only scored lower than the other activities. However, students giving each other feedback, lecturers providing sample texts, and students reading specialized papers and articles all achieved relatively high cumulative GPA (Grade Point Average) and are also valuable in instructional scaffolding. Results. Students believe that teachers need to help them develop skills such as: skills to develop ideas for articles; skills to build article structure; language skills; paragraph writing skills; Sentence linking skills. The lack of these skills explains why, while the self-assessments of children's progress were quite positive for each sub-skill, they were the lowest in the overall assessment of progress in writing and speaking. Practical significance. The role of scaffolding in the writing skills of English majors has been studied by many universities around the world so far. In Vietnam in general and the University of Khanh Hoa, this activity plays an important role in improving students' writing capacity to meet social needs.

No abstract available

No abstract available

Abstract Evidence indicates that graphic organizers, such as concept maps and argument maps, could help students develop a stronger argumentation ability and build effective argumentative structures. In this study, we designed a computer-aided argumentative essay writing system to assist students in learning structures of argumentation and in enhancing their argumentation abilities. In this study, an experiment was conducted on students in an elementary school in New Taipei City for fourteenth weeks. This included a total of 11 classes (272 students), which were divided into three groups: the argument map writing group, the concept map writing group, and the conventional argumentative writing group. The experimental results showed that the argumentative essays produced by the students in the argument map writing group were superior to those written by students in the other two groups.

used graphic organizers and other scaffolding strategies to support reading comprehension and writing of college students. Results show that graphic organizers, as scaffolding tools in the classroom, can

No abstract available

No abstract available

Large Language Models (LLMs) are increasingly being adapted to interpret the physical world through sensor data. However, feeding raw sensor data (e.g., eye-tracking) into these models as text prompts can lead to excessive token overhead and degraded model performance. This study proposes an alternative approach: transforming eye-tracking data into visual representations that serve as prompts for Multimodal Large Language Models (MLLMs). Specifically, we explore three types of visualizations—time-series plot, scanpath, and heatmap—and evaluate their effectiveness on a six-class eye-tracking classification task under zero-shot and one-shot conditions. Our results show that visual prompts not only reduce input tokens by over 85% but also significantly improve accuracy. The heatmap's one-shot accuracy of 73.9% was substantially higher than the 37.8% achieved with raw text. These findings highlight how visual abstraction facilitates the integration of trajectory-based sensor data into MLLM-driven reasoning pipelines.

No abstract available

No abstract available

No abstract available

Current image generation and editing methods primarily process textual prompts as direct inputs without explicit reasoning about visual composition or operational steps. We present Generation Chain-of-Thought (GoT), a novel paradigm that em-powers a Multimodal Large Language Model (MLLM) to first generate an explicit, structured reasoning chain in natural language—detailing semantic relationships,

Visual Commonsense Reasoning (VCR) is a cognitive task, challenging models to answer visual questions, and to explain the rationale behind their answers. While Large Language Models (LLMs) offer potential for this task, VCR’s complex scenes require specialized approaches to activate their commonsense reasoning abilities, as existing Multimodal LLMs struggle with VCR’s visual events and unique reference tags. To address these challenges, we propose EventLens, which enhances VCR through Event-Aware Pretraining and Cross-modal Linking in Supervised Fine-tuning. First, we introduce a new pretraining stage that emulates human cognitive processes to improve LLM comprehension of complex scenarios. Second, during supervised fine-tuning, we leverage reference tags to explicitly bridge RoI features with text, maintaining semantic integrity across modalities. Additionally, instruct prompts and task-specific adapters help integrate LLMs’ knowledge with new commonsense reasoning. Experimental results demonstrate competitive performance with state-of-the-art methods and ablation studies verify the effectiveness of proposed EventLens.

Social intelligence is vital for effective human-AI interaction. While LLMs demonstrate strong text-based social intelligence, the vision modality remains challenging due to the presence of non-verbal social cues. For example, gaze is the primary conveyor of social attention, yet it cannot be accurately perceived and understood by multimodal LLMs (MLLMs). Therefore, we propose GazeCoT, a pipeline using gaze estimation models to provide MLLMs with the attention of people in images or videos. The gaze information is provided as visual and text prompts compiled into a structured context to support MLLM social reasoning. Benchmark evaluation confirms that GazeCoT enhances MLLMs’ social intelligence by improving gaze perception. A user study in a challenging application involving parent-child interactions demonstrates that GazeCoT improves perceived explainability and trustworthiness by aligning MLLM social perception and social reasoning with human norms. We hope that GazeCoT, a versatile plug-and-play pipeline, can enable socially aware, MLLM-based HCI applications.

Scene graphs provide structured visual knowledge that can enhance multimodal large language models (MLLMs) for tasks like visual question answering and image captioning. However, existing approaches inject scene graphs as deterministic, hard prompts, ignoring the inherent uncertainty in visual perception, leading to overconfident and sometimes hallucinated outputs. We propose Probabilistic Scene Graph Prompting (PSGP), a framework that models scene graph generation as a distribution over plausible graphs and encodes this uncertainty into soft, continuous prompt tokens that condition the MLLM. By propagating perceptual uncertainty from detection to language generation, PSGP produces more accurate, faithful, and better-calibrated responses, especially in ambiguous visual scenarios. Experiments on GQA, Visual Spatial Reasoning, and a new Ambiguous-GQA benchmark show that PSGP outperforms strong baselines: including LLaVA-1.5, BLIP-2, and SG-LLaVA, in accuracy, faithfulness, and calibration, while maintaining computational efficiency. Our work establishes a principled pathway toward uncertainty-aware, structured multimodal intelligence.

No abstract available

No abstract available

Flutter is a framework for making mobile applications cross-platform made by Google. From 2019 to 2021 the popularity of flutter is increasing. Flutter use declarative writing style to create layouts. This makes the layout in flutter immutable, and a light blueprint. This research proposes a construct on scaffold concept map method to help students understand the concept of widget arrangement in flutter layout. Construct on scaffold will provide the learner with a framework from an incomplete expert concept map. Some of the nodes and connecting relationships in the framework have been removed, so students must fill in the missing parts with several available answer choices to complete the concept map. To prove the impact of the application of this method, the study was conducted using a pre-post-test group experimental design. Students will do a pre-test, use the EasyFlutter application, and post-test. The results of the pre-test and post-test obtained were tested for normality first, then tested to find out whether there was an average difference between the pre-test and post-test scores. The results of the normality test show that the pre-test data are not normally distributed, and the post-test data are normally distributed, so the next test will use a non-parametric test, namely the Wilcoxon test. The test results show that the post-test mean score is higher than the pre-test mean. Wilcoxon test results also show that the Asymp value. Sig. (2-tailed) of 0.01, so it can be concluded that there is a significant difference between the pre-test and post-test scores. The application of the construct on scaffold method has a significant positive impact on the post-test scores of students related to the concept of widget arrangement in flutter.

While concept mapping is a valuable tool used to organize and represent ideas, the complexity and variability in how concepts are understood, connected, and represented make learners less receptive to using concept maps. Hence, it is important to employ scaffolding and motivational strategies to help students deal with complexity in working with concept maps. Self-efficacy and the perception of the helpfulness of concept map may enhance students’ responsiveness and adaptability of concept mapping. In this quasi-experimental study, we conducted a one-way multivariate ANOVA to examine the difference between the two scaffold conditions (concepts plus labels and concepts only). The study also reports the dimensions of perceived helpfulness of concept map students reported. Results revealed that concept plus labels scaffolding is as effective as concept-only scaffolding in promoting learning among chemistry students. A significant number of students reported that the mapping activity is helpful. Theoretical and practical implications are discussed.  

No abstract available

Structured query language (SQL) is difficult to master because the execution process of SQL statements is invisible. When learning to construct an SQL query, learners must visualise the evolution process of the intermediate datasets of the SQL statement in working memory, which may burden learners’ cognitive load and consequently jeopardise learning outcomes. This study describes the execution process of SQL statements by using concept maps to improve learners’ understanding of SQL. An empirical experiment was conducted using two database courses, namely concept map–based and conventional instruction, to examine the relationship between concept maps and the understanding of SQL from a cognitive load theory perspective. The experimental results demonstrated the superiority of concept map–based instruction over conventional instruction because concept map–based instruction reduces extraneous load but increases germane load. Concept map construction facilitated learner engagement and promoted meaningful learning. Studying the instructors’ concept maps helped learners follow the cognitive structures used by instructors to perform SQL queries, and enabled them to perceive the execution process of SQL queries relatively easily. These results potentially help educators understand the learning difficulties caused by the declarative nature of SQL and motivating researchers to resolve the inherent problem by considering learners’ cognitive processes.

Abstract The rapid development of generative artificial intelligence has created new opportunities for innovation in teacher education, dramatically changing how pre-service teachers acquire new skills and knowledge. In the information age, pre-service teachers’ information technology-enhanced instructional design competence has received widespread attention. This study introduced the generative artificial intelligence represented by ChatGPT to innovate the traditional teaching mode of pre-service education. Learners use ChatGPT to automatically generate courseware and revise courseware according to the framework provided by concept maps. To promote the depth of thinking and motivation of pre-service teachers in the revision process, we proposed an integrated approach with the peer assessment and ChatGPT. The researchers conducted a quasi-experimental study at a university to investigate the effectiveness of this learning strategy. One class of pre-service teachers (N = 46) was the experimental group using peer assessment-based concept mapping-supported ChatGPT generated content (PA-CGPT), and the other class (N = 45) was the control group using conventional concept mapping-supported ChatGPT generated content (C-CGPT). The study suggests that students in the experimental group exhibited significant improvements in practical skills, learning motivation, higher-order thinking tendencies, satisfaction, and reduced cognitive load. Furthermore, in-depth interviews were conducted with both groups of students to examine their learning outcomes and discuss implications for future research.

No abstract available

This experimental research aimed to 1) compare academic achievement in grammatical knowledge between an experimental group taught with the mind map concept and a control group taught with inductive concept and 2) compare the experimental and control group participants’ satisfaction with the management of instructional activities to develop grammatical knowledge after being taught through the mind map concept. A total of 40 Grade 8 Thai students attending an English class in the second semester of the 2020 academic year were recruited through a simple random sampling approach. The sample was divided into two groups, with 20 students in the experimental group and 20 in thecontrol group. The research instruments consisted of 1) a mind map lesson plan; 2) an inductive lesson plan; 3) an academic achievement test; and 4) a satisfaction questionnaire. Descriptive statistics (mean scores and standard deviations) and inferential statistics (independent samples t-tests) were employed to analyze the quantitative data collected in this study.The findings demonstrate that the experimental group participants improved their academicachievement in grammatical knowledge more than the control group participants at the .01 significance level. In addition, the experimental group participants were significantly more satisfiedwith the management of the instructional activities to developgrammatical knowledge than the control group participants at the .01 significance level.

OBJECTIVES Concept mapping is used to promote critical thinking and meaningful learning in health professions education. This study aimed to evaluate dental undergraduates' perceptions of concept mapping as a learning tool for oral radiographic interpretation competency. METHODS A total of 39 third-year undergraduates at a dental college participated in the oral radiological interpretation session. The students received one hour of instruction before creating their concept map (CM). Participants were provided an assignment of four relevant clinical cases to draw an individualized CM. The students' draft and final CM were submitted. The instructor evaluated the CMs to identify whether students could analyze, connect, and organize the information in a meaningful way. Among them, 37 completed the questionnaire about their experiences and responses to concept mapping. RESULTS Students perceived concept mapping positively. The instructor's feedback helped them recognize their misconceptions and fostered greater motivation to learn. Students attempted to integrate basic biomedical knowledge and clinical features through cross-linking. However, they also expressed negative attitudes toward concept mapping regarding time consumption and heavy burden. CONCLUSIONS Concept mapping motivated students' directed learning and gave them the opportunity to recognize their misconceptions. This study suggests the potential use of CMs as an adjunctive learning strategy in the oral radiographic interpretation class.

No abstract available

Electrochemistry, a key topic in introductory chemistry, is a topic that many students struggle with conceptually. Visual learning tools like concept maps allow the externalization of key concepts and the relationships between these concepts may aid in students' acquisition and retention of newly learned electrochemistry concepts. Partnering with a general chemistry instructor, this study explored the effectiveness of two concept map formats—fill‐in‐the‐concepts and student‐corrected maps, on electrochemistry outcomes with undergraduate general chemistry students. To examine students' engagement with these maps, we also explored the role of students' interest. Results indicated that prior knowledge and map condition were significant predictors of students' electrochemistry performance. Specifically, the fill‐in‐the‐concepts condition resulted in better learning outcomes than the map correction condition. Furthermore, students with a higher interest in electrochemistry were more likely to perform well on their respective concept mapping activities, thus, scoring higher on the posttest. These findings suggested that concept maps can effectively represent key concepts of a challenging topic, and are beneficial for students' learning and retention of information. Importantly, instructors may further support learning by providing scaffolds in the form of fill‐in‐the‐concept maps to help students better organize their mental structures and identify misconceptions.

Argumentation is an assessment that identifies HOTs through collaborative learning activities. Learning that supports collaboration through project design is found in project-based learning at the deciding stage. Decisions in planning investigations tend to be less than optimal and the presentation of data at the deciding stage is incomplete so that it is optimized using the instructional concept map technique. The use of concept maps as an instructional technique guides students to find and relate concepts logically so as to stimulate argumentation skills because it explores critical thinking skills through activities of organizing, categorizing, analyzing, and reasoning. The research aims to analyze differences in students' argumentation skills from the use of concept map instructional techniques at the deciding stage – project-based learning. The research design was quasi-experimental. Determination of the sample using cluster random sampling of 70 students who are determined based on the paired F test. The research instrument consisted of an expert concept map and a matter of argumentation. Data on students' argumentation skills were obtained from an assessment that was arranged according to the argumentation component according to Toulmin (2003) is evidence, warrants, backing, qualifier, rebuttal, and claim a total of 11 questions. The data analysis technique used the t-test. The results of the hypothesis test show that the significance value is 0.001 (<0.05), so there are differences in students' argumentation skills from the use of the instructional concept map technique in the deciding – project-based learning stage. The argumentation skills of the students in the experimental class were better than the control class in argumentation components is evidence, rebuttal, and claim. The concept map instructional technique has aspects that can improve students' argumentation skill scores because it helps students focus on finding and connecting concepts based on evidence and data as a result of investigations

The use of concept maps allows learners to embark on a cognitive process of integrating new ideas with their existing knowledge. Thus, helping them to organize, connect and synthesize. This, study focused on the preparation of concept maps based on the perceptions of the teacher-respondents along with the crucial topics in Science 4. The concept maps were integrated in the Alternative Instructional Resource (AIR). Research and Development (R and D) was employed in the study. There were 266 Grade 4 teachers from the Schools Division of Ilocos Norte who were involved in the study. A self-perception survey questionnaire which contains the competencies and topics based on the Most Essential Learning Competencies (MELCs), was used to gather data. The data was analyzed using Statistical mean. Results revealed that the four competencies namely; Matter, Living Things and their Environment, Force, Energy, and Motion and Energy and Space, obtained ratings of Much Appropriate (MA). However, only topics under the four competencies with means higher than four were included in the preparation of concept maps. As regards the validity and acceptability of the material, it was rated Very Highly Valid (VHV) and Very Highly Acceptable (VHA), respectively. It can be inferred based on the results that concept maps are effective tools in enhancing understanding of new concepts and retention of information. Thus, teachers in other subject areas are encouraged to design concept maps to develop critical thinking skills among learners. Keywords: concept maps; alternative instructional resource; most essential learning competencies

No abstract available

Concept mapping, a tool originally developed to facilitate student learning by organizing and visualizing key concepts and their relationships, can also be used to represent the composition of the knowledge contained in a course. In this paper, the authors describe a specific application of concept mapping to help instructors and students visualize the knowledge contained in an introductory programming course of an undergraduate Computer Information Systems program. The authors show how representing the knowledge structure of a course using a concept map can enable the faculty to assess the breadth and depth of the knowledge imparted through the course. The authors discuss how a concept map that depicts the composition of a course can function as a useful instructional tool to assess and improve the quality of instruction that may enable meaningful learning among students.

No abstract available

No abstract available

Knowledge-based visual reasoning remains a daunting task since it not only requires machines to interpret the concepts and relationships from visual scenes but also associate them with external world knowledge to conduct a chain of reasoning on open-world questions. Previous works, however, treat visual perception and language-based reasoning as two independent modules, failing to attend to both modules throughout all stages of reasoning. To this end, we propose Visual Chain-of-thought Prompting (VCTP) for knowledge-based reasoning, which involves the interaction between visual content and natural language in an iterative step-by-step reasoning manner. VCTP contains three stages, see, think, and confirm. The see stage scans the image and grounds the visual concept candidates with a visual perception model. The think stage adopts a pre-trained large language model (LLM) to attend to key visual concepts from natural language questions adaptively. It then transforms key visual context into text context for prompting with a visual captioning model, and adopts the LLM to generate the answer. The confirm stage further uses the LLM to generate the supporting rationale to the answer, which is then passed through a cross-modality classifier to verify that it’s consistent with the visual context. We iterate through the think-confirm stages to ensure the verified rationale is consistent with the answer. We conduct experiments on a range of knowledge-based visual reasoning datasets. We found our VCTP enjoys several benefits, 1). it achieves better performance than the previous few-shot learning baselines; 2). it enjoys the total transparency and trustworthiness of the whole reasoning process by providing rationales for each reasoning step; 3). it is computation-efficient compared with other fine-tuning baselines. Our code is available at https://github.com/UMass-Foundation-Model/VisualCoT.git

Medical Visual Question Answering (MedVQA) lies at the intersection of computer vision, natural language processing, and clinical decision-making, aiming to generate accurate responses from medical images paired with complex inquiries. Despite recent advances in vision–language models (VLMs), their use in healthcare remains limited by a lack of interpretability and a tendency to produce direct, unexplainable outputs. This opacity undermines their reliability in medical settings, where transparency and justification are critically important. To address this limitation, we propose a zero-shot chain-of-thought prompting framework that guides VLMs to perform multi-step reasoning before arriving at an answer. By encouraging the model to break down the problem, analyze both visual and contextual cues, and construct a stepwise explanation, the approach makes the reasoning process explicit and clinically meaningful. We evaluate the framework on the PMC-VQA benchmark, which includes authentic radiological images and expert-level prompts. In a comparative analysis of three leading VLMs, Gemini 2.5 Pro achieved the highest accuracy (72.48%), followed by Claude 3.5 Sonnet (69.00%) and GPT-4o Mini (67.33%). The results demonstrate that chain-of-thought prompting significantly improves both reasoning transparency and performance in MedVQA tasks.

Remote sensing visual question answering (RSVQA) involves interpreting complex geospatial information captured by satellite imagery to answer natural language questions, making it a vital tool for observing and analyzing Earth’s surface without direct contact. Although numerous studies have addressed RSVQA, most have focused primarily on answer accuracy, often overlooking the underlying reasoning capabilities required to interpret spatial and contextual cues in satellite imagery. To address this gap, this study presents a comprehensive evaluation of four large multimodal models (LMMs) as follows: GPT-4o, Grok 3, Gemini 2.5 Pro, and Claude 3.7 Sonnet. We used a curated subset of the EarthVQA dataset consisting of 100 rural images with 29 question–answer pairs each and 100 urban images with 42 pairs each. We developed the following three task-specific frameworks: (1) Zero-GeoVision, which employs zero-shot prompting with problem-specific prompts that elicit direct answers from the pretrained knowledge base without fine-tuning; (2) CoT-GeoReason, which enhances the knowledge base with chain-of-thought prompting, guiding it through explicit steps of feature detection, spatial analysis, and answer synthesis; and (3) Self-GeoSense, which extends this approach by stochastically decoding five independent reasoning chains for each remote sensing question. Rather than merging these chains, it counts the final answers, selects the majority choice, and returns a single complete reasoning chain whose conclusion aligns with that majority. Additionally, we designed the Geo-Judge framework to employ a two-stage evaluation process. In Stage 1, a GPT-4o-mini-based LMM judge assesses reasoning coherence and answer correctness using the input image, task type, reasoning steps, generated model answer, and ground truth. In Stage 2, blinded human experts independently review the LMM’s reasoning and answer, providing unbiased validation through careful reassessment. Focusing on Self-GeoSense with Grok 3, this framework achieves superior performance with 94.69% accuracy in Basic Judging, 93.18% in Basic Counting, 89.42% in Reasoning-Based Judging, 83.29% in Reasoning-Based Counting, 77.64% in Object Situation Analysis, and 65.29% in Comprehensive Analysis, alongside RMSE values of 0.9102 in Basic Counting and 1.0551 in Reasoning-Based Counting.

Large language models (LLMs) empowered by chain-of-thought (CoT) prompting have yielded remarkable prowess in reasoning tasks. Nevertheless, current methods predominantly lean on handcrafted or task-specific demonstrations, lack reliable knowledge basis and thus struggle for trustworthy responses in an automated pattern. While recent works endeavor to improve upon one certain aspect, they ignore the importance and necessity of establishing an integrated and interpretable reasoning system. To address these drawbacks and provide a universal solution, we propose AuRoRA: a one-for-all platform for augmented reasoning and refining based on CoT prompting that excels in adaptability, reliability, integrity, and interpretability. The system exhibits superior performances across six reasoning tasks and offers real-time visual analysis, which has pivotal academic and application value in the era of LLMs. The AuRoRA platform is available at https://huggingface.co/spaces/Anni123/AuRoRA.

No abstract available

Large Language Models (LLM) can significantly enhance the Vision-Language Model's prompting capabilities (e.g. CLIP) by generating detailed and comprehensive prompts. However, the LLMs are prone to generating hallucinated text prompts, leading to misalignment between visual and textual representations in the semantic space. Moreover, sparse sampling of images in low-shot scenarios often leads to incomplete or biased concept representations between two modalities, as the limited data struggles to capture the full range of concept attributes. To this end, we propose Chain-of-Thought Guided Low-shot Debiasing (CoLD), which addresses two key challenges: the textual biases generated by the hallucination issues and the visual biases due to insufficient discriminative features. Specifically, a multi-granularity chain-of-thought (CoT) prompting strategy is proposed that reduces the negative impact of irrelevant textual information, thereby effectively mitigating hallucinations and alleviating the issue of textual bias. Additionally, we propose visual anchoring that utilizes CoT prompts with class attributes and characteristics to generate auxiliary visual features. The anchors optimize the existing semantic space by introducing additional visually-grounded concepts, thereby mitigating visual biases and enhancing low-shot performance. Extensive experiments demonstrate the effectiveness of the proposed method, showcasing notable performance improvements across various datasets.

The emergence of complex reasoning abilities in large language models (LLMs) has sparked great interest, and a variety of prompting techniques was proposed to coax them into emulating human thought processes. In this work, we introduce Think Node-by-Node , a graph-based reasoning framework inspired by mind maps, flowcharts, and other visual aids that help humans tackle complex problems. Rather than generating images directly, our approach leverages standard graph-building and rendering libraries, and requires no fine-tuning, only the model’s native coding capabilities. We further explore a “Fast Thinking” regime, in which a graph-reasoning example provided in the prompt, but the model generates the answers directly, without the full thought process reconstruction. Surprisingly, this approach leads to significant improvement upon baseline in general-knowledge tasks. Remarkably, Think Node-by-Node maintains strong performance even under a strict 25-token budget for answer generation. Across two instruction-tuned LLMs (0.5B and 7B parameters), our Fast-TNbN strategy outperforms baseline prompting techniques, improving accuracy by up to 10%, and exceeds the capabilities of other structured prompting methods under equivalent generation constraints.

Surgical triplet recognition aims to recognize surgical activities as triplets (i.e.,), which provides fine-grained information essential for surgical scene understanding. Existing methods for surgical triplet recognition rely on compositional methods that recognize the instrument, verb, and target simultaneously. In contrast, our method, called chain-of-look prompting, casts the problem of surgical triplet recognition as visual prompt generation from large-scale vision-language (VL) models, and explicitly decomposes the task into a series of video reasoning processes. Chain-of-Look prompting is inspired by: (1) the chain-of-thought prompting in natural language processing, which divides a problem into a sequence of intermediate reasoning steps; (2) the inter-dependency between motion and visual appearance in the human vision system. Since surgical activities are conveyed by the actions of physicians, we regard the verbs as the carrier of semantics in surgical endoscopic videos. Additionally, we utilize the BioMed large language model to calibrate the generated visual prompt features for surgical scenarios. Our approach captures the visual reasoning processes underlying surgical activities and achieves better performance compared to the state-of-the-art methods on the largest surgical triplet recognition dataset, CholecT50. The code is available at https://github.com/southnx/CoLSurgical.

No abstract available

Chain-of-Thought (CoT) prompting has been shown to be effective in guiding Large Language Models (LLMs) to decompose complex tasks into multiple intermediate steps, and constructing a rational reasoning chain for inferring answers. However, the linear nature of CoT falls short from enabling LLMs to effectively handle graph structures, which are essential for personalised recommendation tasks that rely on user-item interaction graphs. To bridge this gap, we propose GollaRec, which leverages a new Graph-of-Thought (GoT) prompting technique in a Multi-modal LLM, namely LLaVA, to effectively exploit the complex structure of the interaction graphs. GollaRec enhances the recommendation effectiveness by integrating both visual and textual "thoughts" into a graph-structured prompt, using both item images and descriptions to produce richer multi-modal user/item representations. In our proposed approach, GollaRec leverages text-graph alignment and graph instruction tuning to allow the Multi-modal LLM to capture complex graph structures. In addition, GollaRec leverages a graph adaptor to integrate user-item interactions into the resulting user/item embeddings, therefore effectively adapting the model to the recommendation task. Our extensive experiments on 6 benchmark datasets demonstrate the superiority of our proposed GollaRec model over 12 existing state-of-the-art models in various multi-modal recommendation tasks, including general and multi-domain recommendation tasks.

No abstract available

With the development of Large models (LMs), recent methods tend to leverage them to complete various downstream tasks, like VQA and image caption. A typical method is Chain-of-Thought (CoT) prompting, which improves the reasoning abilities of LMs by providing intermediate steps. However, existing CoT prompts have two main drawbacks: 1) They typically represent CoT through multiple sentences, which often introduce irrelevant textual or visual context that may confuse LMs. 2) The current CoT methods fail to consider how the contributions of different tokens vary for answer inference. For this, we propose the Multi-Modal Token-of-Thought (MMToT), a novel token-level prompt method to improve LMs’ multi-modal reasoning capabilities. Furthermore, MMToT stress on two strengths against CoT: 1) To prevent LMs from being affected by irrelevant contexts, we propose to extract explicit multi-modal tokens rather than sentences to construct MMToT, enhancing the reliability of generated answers. 2) To ensure that LMs prioritize tokens with high contribution scores during answer generation, we propose a confident decision-making module to evaluate and integrate each token’s contribution in MMToT. Compared to existing methods, the proposed MMToT demonstrates superior performance on Science-QA, MATH, OKVQA, and VQA-introspect datasets. Furthermore, ablation studies and visualization results validate the effectiveness and interpretability of MMToT.

Multimodal Aspect-based Sentiment Analysis aims at identifying specific aspects like celebrities, locations, and organizations, along with their corresponding sentiments. Typically, existing methods comprehend each aspect within images and their relationship without external knowledge. To this end, we propose an Aspect-Attentioned Prompting Framework (AAPF) with Vision-language Prompting (VP) and Attention Aggregation (AA) modules to leverage the fine-grained visual semantics corresponding to the conditional sentences. More specifically, the VP constructs aspect-oriented questions from the perspectives of objects, relations, and sentence-oriented questions, leveraging chain-of-thought prompting to guide the reasoning process. Given the questions, sentences, and images, a visual language model is exploited to perceive the aspect-sensitive intricate semantics, which are further filtered by CLIP scores and condensed into an aspect-relevant summarization. Furthermore, the AA module conducts complementary cross-attention on the summarizations and the conditional sentences to model their interactions by achieving aspect-attentioned features, which can be used to make sentiment predictions. Experiments conducted on two public datasets indicate that our approach achieves superior performance against the state-of-the-art methods.

With the development of Large models (LMs), recent methods tend to leverage them to complete various downstream tasks, like VQA and image caption. A typical method is Chain-of-Thought (CoT) prompting, which improves the reasoning abilities of LMs by providing intermediate steps. However, existing CoT prompts have two main drawbacks: 1) They typically represent CoT through multiple sentences, which often introduce irrelevant textual or visual context that may confuse LMs. 2) The current CoT methods fail to consider how the contributions of different tokens vary for answer inference. For this, we propose the Multi-Modal Token-of-Thought (MMToT), a novel token-level prompt method to improve LMs’ multi-modal reasoning capabilities. Furthermore, MMToT stress on two strengths against CoT: 1) To prevent LMs from being affected by irrelevant contexts, we propose to extract explicit multi-modal tokens rather than sentences to construct MMToT, enhancing the reliability of generated answers. 2) To ensure that LMs prioritize tokens with high contribution scores during answer generation, we propose a confident decision-making module to evaluate and integrate each token’s contribution in MMToT. Compared to existing methods, the proposed MMToT demonstrates superior performance on Science-QA, MATH, OKVQA, and VQA-introspect datasets. Furthermore, ablation studies and visualization results validate the effectiveness and interpretability of MMToT.

No abstract available

This paper introduces Viat-map, a technology-enhanced learning tool designed to improve EFL learners’ reading comprehension by guiding them in reconstructing argumentative structures using Toulmin’s model. Viat-map’s closed-ended, visual interface scaffolds learners through claim, ground, and warrant identification, promoting active engagement and reflection. Experimental results demonstrate that learners using Viat-map significantly outperform those using traditional summarization methods in both immediate and delayed comprehension assessments. Log data analysis further reveals that iterative feedback cycles enhance learners’ accuracy, while behavioral profiles show varied engagement patterns linked to learning outcomes. The latest AI-enhanced Viat-map incorporates adaptive learning pathways and gamification, enabling personalized pacing and richer insights into learner processes. Together, these findings highlight Viat-map’s potential as a data-informed, defensible platform for fostering critical thinking and argumentation skills in multilingual educational contexts.

This is a study of Japanese learners’ perspectives on visual and audio scaffolds in the teaching of second language English reading. We compare the approaches of ER, according to Day & Bamford’s (1998) ten principles (outlined below) in class, and Assisted Repeated Reading (ARR) of a work of fiction to a class. ARR refers to the reading aloud of a text multiple times to students as they follow along silently (see Taguchi & Gorsuch, 2002). Two classes of students in required English classes undertook both ER and ARR. Both the ER and ARR enhanced the students’ comprehension of the texts; in the case of ER, most students chose picture books, which aided their comprehension. In the case of ARR, the embodied presence of the reader, and the modeling of prosody facilitated comprehension.

No abstract available

Reading academic papers is a routine yet cognitively demanding task for researchers and students. The reading process is often fragmented across sessions and shaped by evolving goals, yet most tools treat it as a single-pass, intention-agnostic activity. We present ReadFlect, an interactive system that scaffolds intent-driven, multi-session, and reflective reading of academic literature. ReadFlect enables users to define and manage reading intentions, annotate papers in context, and externalize insights through a diagramming canvas. It further supports reflection through visual analytics that capture reading behaviors over time. We demonstrate how ReadFlect supports more structured engagement with scientific literature and discuss its potential for enhancing comprehension, facilitating synthesis, and supporting collaborative and pedagogical use cases.

Scientific literacy is a cornerstone of modern education, yet its development is hindered by learners' limited mastery of discipline-specific vocabulary. In the Philippine context, this challenge is evident among junior high school students, where decoding, spelling, and comprehension difficulties restrict higher-order scientific thinking. This study investigates the effectiveness of the "Who Am I?" intervention in improving Grade 7 learners' mastery of science terminologies at San Martin National High School. Initial diagnostic results from the Philippine Informal Reading Inventory (Phil-IRI) revealed that 26 out of 52 learners were at the Frustration level, showing severe challenges in reading and spelling scientific vocabulary. These difficulties impeded comprehension and reduced confidence, underscoring the need for targeted interventions aligned with the MATATAG curriculum. A three-month intervention was implemented, integrating clue-based identification, spelling reinforcement, visual scaffolds, peer tutoring, and technology-assisted instruction. Results showed significant improvement, with Mean Percentage Scores rising from 42 percent (Poor Mastery) to 78 percent (Proficient Mastery). A paired t-test ($t = 15.62, p = 0.0001$) confirmed statistical significance. Qualitative findings revealed enhanced confidence, clearer pronunciation, improved spelling, and greater participation in class discussions. Learners found the intervention "fun" and less intimidating, while teachers observed increased engagement. The study supports multimodal and contextualized vocabulary instruction in science education. By embedding literacy-focused strategies, the "Who Am I?" intervention demonstrates that low-cost, replicable approaches can strengthen foundational skills in resource-constrained classrooms. Findings highlight benefits for learners, teachers, curriculum developers, and policymakers, offering evidence for scalable interventions. 

No abstract available

The team at Pop Culture Classroom makes a great case for using comics in literacy education, pointing out the various ways this visual storytelling style scaffolds reading comprehension.

Reading comprehension is a critical component of academic achievement, yet many English as a Foreign Language (EFL) learners face challenges in processing complex texts. This study investigates the impact of visual information-based reading materials on improving comprehension among Indonesian EFL university students. A mixed-methods, cross-sectional design was employed over three academic semesters. A total of 228 students were assigned to either an experimental group, which received reading materials incorporating visual elements, or a control group, which received text-only materials. Quantitative data were collected using standardized comprehension assessments, while qualitative insights were gathered through student reflections and interviews. The experimental group demonstrated significantly higher performance in literal comprehension compared to the control group (p 0.05, Cohen’s d = 0.81), indicating a large effect size. Qualitative data reinforced these findings, revealing that visual aids enhanced student engagement, understanding, and information retention. The integration of visual elements into reading materials effectively supports comprehension in EFL contexts. Visual information appears to reduce cognitive load and facilitate meaning-making, particularly in learners processing texts in a non-native language. Visual-enhanced reading materials significantly improve literal comprehension among Indonesian EFL university students. These findings underscore the importance of incorporating visual literacy strategies in EFL pedagogy and call for further research into interactive and digital visual tools to support reading comprehension.

Visual literacy is important in understanding a language text. One of the characteristics that might improve students' comprehension of reading materials is visual literacy, which includes the capacity to comprehend and interpret visual features like pictures, graphs, and diagrams in text. The use of visual literacy media has become an increasingly important topic in education, especially in the context of students' reading comprehension at the junior high school level. This article aims to find out how visual literacy affects 8th grade reading comprehension at SMPN 01 Buay Madang. This research method uses qualitative methods by using a qualitative descriptive approach and data collection tools in the form of observation, and documentation. The results showed that: (1) there is a significant difference between students who use visual literacy media and those who do not; (2) visual literacy media can help students understand texts more easily and quickly because of the visual elements included in the text that explain the information; and (3) students who use visual literacy media have better comprehension skills and are more adept at applying information when compared to students who only use text without any visual elements. This suggests that visual literacy can be an effective strategy to increase comprehension, especially in digital learning contexts where visual aids are more prevalent. This study offers recommendations for a curriculum that focuses more on the use of visual literacy in classroom instruction with the aim of helping students understand the material more thoroughly and raising the standard of instruction in schools.

No abstract available

The complexity of physiological processes in the digestive system often poses significant cognitive challenges for junior high school students, leading to suboptimal conceptual mastery. Conventional learning materials often lack the visual scaffolding needed to help students organize abstract concepts. This study aims to develop Visual Mind Map Worksheets integrated with the Discovery Learning model and to analyze their validity, practicality, and effectiveness in improving students' understanding of science concepts. This research utilized a Research and Development (R&D) design following the 4-D model (Define, Design, Develop, and Disseminate). The trial sample consisted of 28 eighth-grade students at Asa Cendekia Junior High School, selected using purposive sampling. Data were collected using expert validation sheets, learning implementation observation rubrics, and a written test measuring cognitive levels C2 (Understanding), C3 (Applying), and C4 (Analyzing). The data were analyzed using descriptive statistics, paired sample t-tests, and N-gain scores. The results indicated that the developed product met the criteria for high feasibility. Expert validation yielded an average score of 3.90 (Very Valid), while the practicality test showed a 98% implementation rate of learning activities (Very Practical). Statistical analysis revealed a significant difference between pretest and posttest scores (p = 0.000 < 0.05). Furthermore, the N-gain analysis demonstrated an improvement in the "Medium" category across all indicators, with a hierarchical achievement pattern: Understanding (0.635), Applying (0.416), and Analyzing (0.444). These findings conclude that the Visual Mind Map Worksheet is a feasible and effective instructional tool. By leveraging visual organization, the worksheet acts as a cognitive scaffold that helps students reconstruct complex biological information into a coherent understanding. Keywords: visual mind map-based worksheets, discovery learning, science concept understanding, digestive system.

This study, based on the Cognitive Theory of Multimedia Learning, explores the effects of different prompting modes (visual and text-based) on learners’ computational thinking and problem-solving abilities. A Blockly-based visual block learning system was developed with a robot vacuum task to guide students through problem-solving activities. A quasi-experimental design was used, including 62 university and graduate students divided into a visual-prompt group and a text-prompt group. The results showed a significant difference in computational thinking performance in the post-test, with the visual-prompt group outperforming the text-prompt group. In terms of problem-solving abilities, the visual-prompt group showed significant improvement in problem-solving confidence and personal control. Qualitative analysis also indicated that visual prompts helped students decompose problems, plan strategies, reduce cognitive load, and improve problem-solving performance. Overall, the findings suggest that visual prompts can enhance students’ computational thinking, problem-solving abilities, confidence, and initiative, providing useful insights for future research and instructional design.

This study investigates the impact of AI-based scaffolding tools on reading comprehension and cognitive strategy use among second language (L2) learners with lower working memory capacity. Specifically, it compares the effectiveness of adaptive, visual, and interactive scaffolding in enhancing reading proficiency and fostering strategic reading behaviors. Participants were randomly assigned to one of three scaffolding conditions or a control group. Pre- and post-intervention assessments measured changes in reading comprehension scores and self-reported strategy use via the Survey of Reading Strategies (SORS). Results indicated that learners in the adaptive scaffolding condition demonstrated significant improvements in both reading comprehension and the adoption of higher-order strategies such as predicting and summarizing. Qualitative feedback highlighted increased confidence and reduced cognitive load. These findings suggest that AI-based adaptive scaffolding can effectively support L2 learners with lower working memory by providing personalized, just-in-time assistance, thereby promoting deeper engagement with texts and more autonomous learning habits. Theoretical implications for cognitive load theory and practical applications for technology-enhanced language instruction are discussed.

This study aims to examine the effectiveness of a visual-scaffolding-based Nahwu formula chart intervention on students’ grammatical ability. Using a quantitative, quasi-experimental design (One-Group Pretest-Posttest Design), this study involved 30 students at Fatkhul Mu’in Ali Masykur Islamic Boarding School in Wonosobo as research subjects. Data collection techniques included tests, treatment, observation, and documentation. The results of the study show a significant increase in the average scores. The lowest student score improved from 38 to 63, while the highest score reached 96. The effectiveness analysis using the N-Gain score yielded a moderate value. These findings confirm that the use of formula charts as a visual instrument can reduce students’ cognitive load and facilitate pattern recognition in Arabic syntactic analysis. This study recommends the integration of visual media into the ilmu alat curriculum in Islamic boarding schools to accommodate the learning styles of contemporary students.

The growing importance of 3D animations in current teaching approaches becomes increasingly apparent, offering an effective way to visualize complex spatial concepts and processes in geography learning through outstanding visual representation and details. However, the effects of detail richness and text load of 3D animation on learning about processes remain unclear. Addressing this research gap, the present study adopts a quasi-experimental design involving four classes (n = 106) in the context of a geography lesson and evaluates four conditions in a 2 × 2 between-subjects design consisting of detail richness (high vs. low) and text load (high vs. low). The lessons on the rotation and revolution of the Earth were delivered by the same instructor across all conditions. Knowledge acquisition, cognitive load, learning experience, and emotions of students were measured. The results revealed that students were significantly better able to acquire knowledge immediately when exposed to the high detailed visuals but low text load condition. Low detail richness and high text load independently resulted in increased cognitive load. We also observed a significant effect of detail richness on the dimensions of pleasure and arousal, with higher levels of details associated with larger values in these dimensions. This research suggests that when the learning objective necessitates the engagement of mental imagery, incorporating detailed visuals can facilitate learning. The findings contribute to our understanding of how detailed imagery is linked to learning objectives about processes and expand our knowledge regarding the design of detail richness and text of 3D animation in the context of geography learning.

Mathematical word problem-solving remains a significant challenge for early primary students due to the abstract nature of linguistic information and cognitive load constraints. This study investigates the effectiveness of the Sketching Technique in enhancing the mathematical achievement of Year 2 students, with a specific focus on word-problem solving involving basic arithmetic operations. A quasi-experimental research design was employed involving 60 Year 2 students from a primary school in the Northeast District of Penang. The participants were divided into a treatment group (n=30) that used the Sketching Technique and a control group (n=30) that followed conventional teaching methods. The research instrument was an achievement test consisting of 10 word-problem questions, with a total score of 10 marks. The test's internal consistency reliability was acceptable (Kuder-Richardson Formula 20 = 0.81). Data were collected using pre-test and post-test instruments and analyzed using paired and independent-samples t-tests. Findings revealed a significant improvement in the treatment group, with the mean score increasing markedly from 2.33 on the pre-test to 9.53 on the post-test. Conversely, the control group showed minimal improvement, with mean scores rising from 2.30 to only 2.90. Statistical analysis confirmed a significant difference between the two groups, with the treatment group demonstrating more consistent performance as evidenced by a reduction in standard deviation from 1.73 to 0.51. Furthermore, the effect size (Cohen's d) was calculated to be 5.65, indicating very large practical significance of the intervention. These results suggest that integrating visual-based pedagogical innovations is crucial for improving numeracy outcomes and fostering critical thinking in early primary education.

This study evaluates the impact of EduFIGURE9, a visual pedagogical tool designed to enhance inclusive learning in fashion illustration, focusing on reducing cognitive load and improving engagement. The research addresses the challenge of bridging beginner skills and industry-ready competencies in Technical and Vocational Education and Training (TVET) contexts, particularly for hearing-impaired learners. We employ a mixed-method approach, combining quantitative surveys and performance metrics from 122 participants, including a hearing-impaired subgroup, with qualitative feedback on usability and accessibility. The tool integrates Visual Learning Theory, User-Centered Design (UCD), and Universal Design for Learning (UDL) principles to provide structured visual scaffolding. Quantitative results demonstrate high usability scores, with learners reporting clear visual guidance and improved task completion rates, while qualitative insights reveal reduced reliance on verbal instruction among deaf students. Lecturers observed increased instructional efficiency, allowing more time for creative exploration. The findings highlight EduFIGURE9’s effectiveness as a low-tech assistive technology, significantly reducing cognitive load and fostering engagement across diverse learner groups. This study contributes to inclusive education by validating a practical, scalable tool that promotes equitable learning opportunities in fashion illustration. The results underscore the pedagogical and technical value of visual scaffolding in specialized vocational training, offering implications for educators and curriculum designers seeking to enhance accessibility and instructional efficiency.

In the epistemological constellation of contemporary visual education, web comics establish themselves as semiotic devices with a dual pedagogical value. On the one hand, they represent the advanced synthesis of multimodal codes (visual, textual, sound); on the other, they structure participatory learning environments typical of digital culture. This contribution starts from the theoretical premise that digital comic narratives, in their hyper-textual and algorithmic essence, constitute true liminal spaces where ontological meanings are negotiated and transformative visual literacy skills are developed. The research, conducted according to the methodological paradigm of design-based research and rooted in the framework of multimodal social semiotics, demonstrates how sequential narrative structure and visual metaphors reduce cognitive load, through scaffolding. The interactive mechanisms typical of the medium (comments, sharing, remixes) promote an aesthetic of participation, transforming students from passive users to producers of knowledge, according to the principles of connected learning. The analysis focuses on: Panda Likes Bevilacqua, Totally Unnecessary Comics by Leone; Rossoni’s Rouge Worms, Lele Corvi’s strips, Natangelo’s cartoons. The limitations of the study lie in the limited sample and its preliminary nature, as it analyses the device itself without evaluating its implications in the classroom or the professional skills (TPACK) necessary for teachers. Large-scale teaching feasibility remains to be investigated in future applied experiments, which involve the direct involvement of classes and teachers. From a pedagogical perspective, web comics are effective teaching tools for students with ASD. Their community-based nature requires a recalibration of traditional pedagogical frameworks towards more ecological approaches.

The development of visual literacy is examined through the lenses of disciplinary discourse and cognitive load, and the need for examining discrete elements of representations independently to reduce the cognitive load of learners is rationalized.

No abstract available

Comics are increasingly recognised as a promising educational medium that integrates visuals and text to convey knowledge in an accessible, engaging and meaningful way. Unlike conventional text-heavy materials, comics provide a multimodal learning experience that draws on the strengths of both linguistic and visual representation. For pupils with learning disabilities (LD), including dyslexia, Attention Deficit Hyperactivity Disorder (ADHD) and autism spectrum conditions, such resources are especially valuable. Traditional instructional texts often create barriers to reading fluency, comprehension and motivation, whereas comics offer structured scaffolding through sequential panels, simplified dialogue and expressive imagery. These features can reduce cognitive load, sustain attention and foster active engagement with content. This concept paper explores the potential of educational comics as a visual medium by examining the global history and evolution of comics as a form of communication, their categorisation into narrative, informative and educational types, and their pedagogical relevance for pupils with LD. Several theoretical perspectives guide this exploration, including Dual Coding Theory, Cognitive Load Theory, Automaticity Theory and Self-Determination Theory. Based on these foundations, a conceptual framework is proposed to illustrate how comics can support learning processes and outcomes. The framework highlights the contributions of comics to reading fluency, comprehension, motivation and socio-emotional development. The discussion also considers potential challenges in adopting comics within special education, particularly issues of teacher preparedness, curricular alignment and the availability of appropriate resources. Nevertheless, the inclusive and motivating nature of comics positions them as valuable pedagogical tools for promoting accessible, equitable and engaging education for diverse pupils. This positions educational comics not merely as supplementary resources but also as transformative tools for inclusive education.

In this Action Research study, we explored how scaffolded instructional strategies enhanced undergraduate English as a Foreign Language (EFL) students’ confidence and engagement in reading English poetry. We conducted this study among twenty bachelor level, first year EFL students in one month-long poetry reading project at a campus in a rural area of Nepal. When we observed students’ struggle with figurative language, unfamiliar vocabulary, and culturally distant imagery in poetry reading lessons, we introduced scaffolded poetry instruction strategies which included guided recitation, collaborative reconstruction of poetic lines, visual and contextual scaffolds, and structured discussion on interpretation. In addition to these interventions, we taught them poetic devices and provided opportunities to compose poems. We collected data from classroom observations, informal interviews with the participants, students’ artefacts, and our reflective accounts. The study found that the scaffolding activities reduced cognitive load and enhanced students’ understanding of the poems. The study also showed that scaffolded instructions have visible improvements in students’ explorations of text, interpretation, and creativity. These findings align with recent evidence that scaffolding strategies, such as peer support, guided noticing, and concept-mapping, significantly improve comprehension and learner agency in EFL contexts. The study demonstrates that scaffolded poetry instruction can transform poetry from an intimidating task into an accessible and meaningful component of language learning in the context, where students get limited opportunities to explore poetry.

As generative AI continues to reshape educational landscapes, prompt-to-image technologies offer new possibilities for enhancing visual pedagogy. This study investigates the integration of NanoBanana a lightweight, prompt-driven image generation tool, into higher education settings to support multimodal learning and cognitive scaffolding. Grounded in Dual Coding Theory and Cognitive Load Theory, the research explores how AI-generated visuals derived from student and instructor prompts can improve comprehension, engagement, and retention in complex subjects such as ICT, Game Studies, and Systems Analysis. Using a mixed-methods approach, the study analyzes student performance data, visual rubric evaluations, and qualitative feedback from learners and educators. Findings reveal that NanoBanana-generated images significantly aid in conceptual clarity, reduce extraneous cognitive load, and foster learner autonomy. The paper proposes a practical framework for integrating prompt-to-image tools into curriculum design and instructional workflows, offering actionable insights for educators seeking to advance AI-enhanced teaching practices in line with SDG4 and the evolving demands of the AI era.

Higher-order functions are increasingly common in modern programming languages, yet there is a shortage of evidence-based tools and teaching strategies to help students learn them effectively. We introduce a visual tool that lets learners construct, view, and execute higher-order functions using direct manipulation and programming by demonstration. To evaluate its effectiveness, we conducted a randomized, within-subjects study with 27 university students, comparing our tool against Python as a control. The results show that students performed significantly better and reported lower cognitive load when solving simple problems with our tool. However, both groups showed similar performance on tasks that involved mapping input-output pairs to the correct higher-order function. Our findings suggest that visual, direct-manipulation tools can help students develop stronger procedural knowledge of higher-order functions, although additional scaffolding may be needed to foster deeper conceptual understanding.