Middle-schoolers often struggle to create research questions and hypotheses that are clear, specific, and empirically testable, especially in resource-limited settings [1, 2]. Evidence suggests that AI tutors deliver real-time, individualized feedback that sharpens scientific reasoning [3, 4]. AI-driven feedback enhances clarity, specificity, and logical coherence of students' inquiries by highlighting gaps in reasoning and fostering independent problem-solving [5].

It has become increasingly important to understand the challenges in IoE-based engineering education due to stress among the academicians who are forced to work remotely due to the lockdown imposed as part of stringent COVID-19 protocol. In this study, an investigation is carried out to understand the nature of stress and its impact on a cohort chosen from academics spread across India. A semi-structured survey investigates the nexus between work & stress among young academicians- how it varies as per demographic, neighborhood, entertainment, lectures, and performance. The study shows that the post-graduate participants and those aged 26–30 years are 14-15% more stressed. Those who are working-from-home in rural areas are more stressed. Participants with the highest levels of stress are found to underrate their own performance as well as others' involvement. It is observed that spending more time on entertainment or developing new interests may reduce stress levels. The study also revealed that carefully designed lectures might lower stress levels as most participants realized. The study contributes to the knowledge base for considering stress associated factors in learning activity.

A Learning Analytics Dashboard is a quick and efficient way for instructors to track the activities of students. In massive online learning scenarios like an international robotics competition, a dashboard is a critical tool for instructors to ensure continuous engagement of participants. Previous research on learning analytics dashboards focused on the effectiveness of dashboards and learning analytics on students along with factors affecting its success. This research discusses a dashboard developed for a massive robotics competition through which each year thousands of students are trained in engineering skills in an online Project Based Learning approach. The dashboard is developed using the dataset for the competition conducted during September 2020 to April 2021 in which more than 10,000 undergraduate students from 572 academic institutions across 7 countries participated. Team characteristics like demographics, feedback, scores, online activity, etc. are considered to cluster teams and develop models to predict the retention of participants. The Machine Learning (ML) model was able to achieve an accuracy of 80.7% and a recall value of 83.9% to identify dropping teams. Clustering provided insights on how these characteristics affected the performance of participants. These predictions along with participant engagement and feedback data was displayed on the dashboard. This visualization helps instructors identify teams requiring guidance or scaffolds to continue participation. Feedback from instructors shows the dashboard to be a promising tool for effectively managing massive online competitions.

MOOCs are experiencing high enrolment statistics and academics are focusing on ways to improve the retention rates in MOOCs. e-Yantra imparts hands-on engineering skills in a highly scalable manner to undergraduate students in engineering, polytechnic and science colleges. The e-Yantra MOOC takes form of a Robotics Competition where students are mentored to solve problems modelled as “Themes.” We study the effect of “difficulty levels” on participant retention of two Themes (hands-on MOOCs). The performance of 800+ students that participated in each of these MOOCs (1600+ total students) offered in 2017 and 2018 is considered. In the light of increasing popularity of online learning especially through MOOCs, this study contributes to the literature of “funnel of participation” in MOOCs teaching conceptually difficult engineering topics in a hands-on Project Based Learning mode. Our study brings contrast with the general comprehension from prior literature that shows MOOCs teaching easier concepts to have high completion rates and MOOCs teaching difficult concepts to have low completion rates. We find simplicity and coherence of MOOC design, regardless of difficulty level of MOOC to have a stronger effect on learning outcomes. Conceptually difficult MOOC in our study has “2.75” times higher completion rate compared to the easier MOOC. This affirms that it is better to teach solving “one challenging problem” than to teach solving a series of easy problems since the latter leads to “conceptual clutter” among participants and as a result a fatigue and consequential increased dropouts.

Software conceptual design knowledge and skills are essential for graduating computer engineering students. Several approaches have been used to teach software conceptual design through various frameworks and learning environments. Our brief literature survey found that none of these systems analysed learner engagement with design elements using learning analytics. This study aims to analyse the learner engagement with design elements as they create software designs. We analysed log data of N=37 students in Think & Link, a technology-enhanced learning environment (TELE), based on the FBS design framework. The differences in student engagement are compared for two categories of learners, high and low, based on the learning gain in the pre and post-test performance. Differences and similarities in logs of actions were identified for high and low learners. Process Models were used to identify differences in the sequence of engagement with various design elements. We find high learners to identify significant design elements and create links among them. The analysis also provided insights on points for scaffolding to help low learners create meaningful designs.