ETR&D

Artificial intelligence (AI) has ushered in an era where Computational Thinking (CT) emerges as a crucial skill. This interconnected, data-driven landscape necessitates deeper CT skills, enabling students to effectively navigate both the opportunities and challenges presented by AI. The flipped classroom model has gained widespread acceptance in CT education owing to its adaptability, interactivity, and personalization. However conventional implementations of the flipped classroom model still face significant limitations such as insufficient student autonomy in learning processes and a need for enhanced quality of classroom interactions. Addressing these, our study introduces a synergy of flipped classrooms with AI, termed the Flipped Classroom with AI Learning Companion teaching model (FC-AIC). This model aims to enhance high school students CT, self-efficacy, and motivational levels compared to traditional flipped classrooms (FC). Our quasi-experimental research involved 60 first-year high school students from central China, divided into control (FC) and experimental (FCAIC) groups over eight weeks. Students completed a questionnaire to self-appraise their CT, self-efficacy, and motivation at pre-intervention and post-intervention stages. Results indicate that FC-AIC significantly boosts student self-perceptions of their own CT skills, specifically in problem-solving and creativity, and improves perceived self-efficacy and intrinsic motivation without notably impacting perceived extrinsic motivation. This investigation highlights how integrating flipped classrooms with AI technologies can synergistically enhance educational paradigms offering new insights about practical innovations for intelligent educational frameworks.

The current study examined the effects of using a commercially available computer program on the early literacy skills of preschool children and the relationship between fidelity and improvement in literacy skills. Parents and their preschool children were recruited to voluntarily participate in an educational app study. Forty-two preschool children were randomly assigned to an intervention group, which used OgStar Early Reader, or a control group, which used IXL Math. The recommendation was to use the program for 15–20 min per day for five days a week over eight weeks in the summer prior to kindergarten. Three Dynamic Indicators of Basic Early Literacy Skills (DIBELS) measures were used to assess literacy skills: Letter Naming Fluency (LNF), Phoneme Segmentation Fluency (PSF), and Nonsense Word Fluency (NWF). A total of 33 children completed posttests. The intervention group scored statistically significantly higher on LNF posttests (g = 0.41, p = .025) and NWF- correct letter sounds posttests (g = 0.52, p = .009) over the control group. No statistically significant differences were found between the groups for PSF (g = 0.19, p = .458) or NWF- words recoded correctly (g = 0.61, p = .057). Overall, children completed between seven and 131 lesson segments during the study, suggesting fidelity to the planned intervention varied across participants. The number of lessons completed was moderately related to participant gains on LNF and NWF (r = 0.31–0.38). These findings provide some initial support that the use of the early reader app may improve alphabetic knowledge for preschool children.

The study investigates the latent profiles of learners based on motivated self-directed learning (MSDL) and collaborative learning (CL) within an e-learning environment, focusing on how peer-to-peer confirmation is linked to learners' engagement and learning outcomes. Grounded in Self-Determination Theory, and drawing on empirical insights from recent benchmark studies, the research identifies two distinct learner profiles: one with low MSDL and CL engagement and the other with high engagement in self-directed and collaborative learning, including ICT-based activities. Logistic regression analysis reveals that peer-to-peer confirmation, particularly individual attention, predicts class membership, though the overall explanatory power is limited. Additionally, moderation analysis demonstrates that while peer assistance can support engagement, excessive support may diminish the benefits of self-directed learning on subjective achievement. The findings underline the importance of adapting social support strategies to different learner profiles, highlighting the motivational function of peer confirmation in fostering autonomy, relatedness, and competence. This study provides insights on how to optimise e-learning environments by balancing external support and learner agency, and offers practical recommendations for digital pedagogy design.

As a part of the Erasmus + funded project Super Cyber Kids, we have devised a skills framework to enhance the protection of children aged between 8 and 13 years in their Internet and online activities. The framework was developed through a systematic literature review, followed by a two-round Delphi study for validation. During this process, we identified N = 40 relevant studies and obtained validation from up to N = 18 professionals in cybersecurity, education, and cybersecurity education during the Delphi study. The resulting framework is presented as a matrix and offers a comprehensive set of skills aligned with the NIST (National Initiative for Cybersecurity Education) cybersecurity framework, specifically tailored to promote cybersecurity awareness among children in this age group.

Virtual scientific inquiry (VSI) environments are critical information technologies for improving the quality of education. However, the issue of students’ cognitive-affective profiles has received little attention, making it difficult to promote both cognitive achievements and positive affective factors. Based on the log-files and self-report data from 569 students, our study proposes a solution that integrates both theoretical underpinnings (including Cognitive Constructivism, Self-efficacy Theory, Flow Theory, Gender Schema Theory, and Control of Variable Strategy) and AI algorithms-driven analyses (i.e., unsupervised machine learning, pattern mining, and statistical analyses) to address this issue. Specifically, our study identifies two profiles: 350 thriving students and 219 struggling students, through the analysis of inquiry achievements, self-efficacy, and flow experience. Moreover, we found no significant gender difference between the two cognitive-affective profiles. Finally, we found that while thriving and struggling students shared similar foundational characteristics, thriving students exhibited more strategic behavioral frequencies and patterns. These findings can advance understanding of theoretical underpinnings and guide targeted support for students in VSI.

Due to the popularity of video lectures, an increasing number of older adults are engaging with this form of learning. Furthermore, co-viewing has become a common type of social interaction. Despite the popularity of video lectures among older adults, there is limited understanding of how to effectively support their learning through this medium. The present study tested the interaction effects of co-viewer similarity (young vs. older co-viewer) and the type of learning content (declarative vs. procedural knowledge) in video lectures on older adults’ learning performance, motivation, cognitive load, and PFC activation as measured by fNIRS. Our results found that older adults showed better learning performance, greater motivation, and reduced cognitive load when co-viewing with an older viewer compared to a young one, regardless of the type of learning content. Interestingly, for declarative knowledge (i.e., Traditional Chinese Medicine, older adults exhibited greater activation in the FEFs, S1, and SMC regions when co-viewing with an older viewer compared with a young viewer. For procedural knowledge (i.e., digital technology), they exhibited less activation in the dlPFC and SMC regions when co-viewing with an older viewer compared with a young viewer. Our findings suggest that educators should consider the age of co-viewers when designing video lectures for older adults, irrespective of the type of knowledge being taught.

Studies on learner–learner and learner–instructor interaction provide insight into the preferences and perceived effects of interaction types. However, evidence of the impact on measured performance resulting from substituting learner–content interaction for learner–learner interaction is sparse. Using an experimental design, this study examined the impact of substituting interaction types on perception of workload, perception of learning, and measured performance in an online, asynchronous, undergraduate-level setting of formal distance education. The results of this study showed (a) learner–learner interactions were perceived to be significantly more work than learner–content interactions, (b) learner–content interactions were perceived to be significantly more helpful in learning the material, (c) there was no significant difference in measured performance between the two interaction types, (d) interaction type did not significantly moderate the relationship between perception of workload and measured performance, and (e) interaction type did significantly moderate the relationship between perception of learning and measured performance. Implications and recommendations are also provided to inform future researchers and practitioners.

When students use critical thinking (CT) skills to identify, place, and link nodes to construct a concept map, what are the effects of constructing concept maps on students’ CT abilities? To what extent do prior studies control what mapping processes and CT skills are demonstrated to students and examine what CT skills students use while constructing concept maps? This review examines the effectiveness of concept mapping compared to traditional teaching methods for enhancing CT abilities. It includes 17 studies involving 1363 students that measured CT skills (e.g., analysis, evaluation, inference, induction, and deduction) using standardized tests in experimental and control groups. Our analysis revealed that the findings on the effects of concept mapping on CT skills are mixed. The studies implemented research designs and CT measures that lacked consistency. They reported minimal information on which CT skills were demonstrated to students, which CT skills students used to construct their maps, and how specific skills impacted the quality of students’ concept maps. This information is necessary to establish consistency and fidelity in implementing concept maps as an instructional method, identify possible variables contributing to the mixed findings, and enhance the effectiveness of concept mapping. Accordingly, this review aims to identify specific methodological gaps that may account for these inconsistencies and guide the refinement of future research design.

This phenomenographic study explores how six leading academic experts in the design and instructional design field experience creativity in instructional design, specifically, how they perceive, define, and use creativity in instructional design. Phenomenography describes the diverse ways these experts experience, conceptualize, and integrate creativity into ID, highlighting the complex nature of creativity within the field. The findings reveal a range of interpretations and applications of creativity, emphasizing the challenge of defining and incorporating creativity explicitly in ID education. Despite recognizing creativity as a critical skill for the twenty-first century, there is still a lack of explicit inclusion in the instructional design curricula. This study contributes to understanding how creativity is perceived in instructional design by presenting four distinct categories of description. By describing these experts' perspectives on the role of creativity in their work, this study contributes to the ongoing discourse on the significance of creativity in instructional design. It highlights the need to reevaluate how creativity is fostered within instructional design programs and advocates for a more integrated and explicit approach to fostering creativity within the classroom to prepare future instructional designers for the complexities of the twenty-first-century learning environment.

Predicting the level of student learning engagement in online learning is crucial for student success, especially for asynchronous courses. While digital traces can track students’ activity on the platform and help to measure the engagement level, they could provide contradictory results, so it is crucial to incorporate complementary methods which can triangulate the findings obtained from digital traces. This study aimed to develop and validate a model to determine the level of learning engagement in adult learners on an asynchronous online platform using a mixed-method approach. Data from digital traces, surveys, and interviews were combined. The study involved 2234 students and employed Extreme Gradient Boosting and Logistic Regression with L2 regularisation models to predict the level of engagement. The Extreme Gradient Boosting model more accurately predicted students in the low engagement group, providing crucial support for potentially vulnerable students. The number of finished homework assignments and attempts were found to increase the probability of high engagement. The diversity of activities, such as access to text materials, played a pivotal role in sustaining engagement. Interviews corroborated these results, suggesting the model effectively reflects engagement levels. The article discusses implications for constructing similar models in future research.