Current initiatives to introduce machine learning (ML) in high schools typically rely on simplified content tthat avoids mathematical foundations, creating a significant gap between secondary and advanced ML education. We present a fundamentally different approach to scaling rigorous ML education: a blended learning framework enabling advanced STEM track students to learn directly from professional ML resources while addressing the critical challenge of teaching capacity limitations. Our framework strategically positions prominent Contemporary Digital Learning Resources (CDLRs) --- high-quality resources created by professionals for adult learners --- at the center of instruction, redistributing educational responsibilities among teachers, students, and digital resources to transform classroom dynamics. Our framework addresses a critical scaling challenge in ML education: the shortage of teachers with ML expertise. By drawing on Cognitive Apprenticeship, Community of Inquiry, and Self-Efficacy theories, we create structured learning environments that potentially extend apprenticeship models beyond direct expert-novice relationships. Initial implementation across six high schools shows promising results, with external evaluation revealing substantial student achievement and high completion rates. Student data indicates a consistent preference for professional resources over simplified alternatives, suggesting the value of authentic disciplinary engagement for such science-track students when adequately supported. Our ongoing work focuses on developing empirical measures to evaluate how effectively digital resources implement cognitive apprenticeship elements and identifying critical factors affecting successful implementation. Our work contributes to theoretical understanding of cognitive apprenticeship in digital environments and provides empirical evidence for redistributing educational roles around professional learning resources, with implications that extend beyond ML education to other rapidly evolving technical fields where traditional teacher preparation may prove insufficient.

Explainable Artificial Intelligence (XAI) seeks to render Artificial Intelligence (AI) models transparent and comprehensible, potentially increasing trust and confidence in AI recommendations. This research explores the realm of XAI within unsupervised educational machine learning, a relatively under-explored topic within Learning Analytics (LA). It introduces an XAI framework designed to elucidate clustering-based personalized recommendations for educators. Our approach involves a two-step validation: computational verification followed by domain-specific evaluation concerning its impact on teachers’ AI acceptance. Through interviews with K-12 educators, we identified key themes in teachers’ attitudes toward the explanations. The main contribution of this paper is a new XAI scheme for unsupervised educational machine-learning decision-support systems. The second is shedding light on the subjective nature of educators’ interpretation of XAI schemes and visualizations.

AI-powered educational technology that is designed to support teachers in providing personalized instruction can enhance their ability to address the needs of individual students, hopefully leading to better learning gains. This paper presents results from a participatory research aimed at co-designing with science teachers a learning analytics tool that will assist them in implementing a personalized pedagogy in blended learning contexts. The development process included three stages. In the first, we interviewed a group of teachers to identify where and how personalized instruction may be integrated into their teaching practices. This yielded a clustering-based personalization strategy. Next, we designed a mock-up of a learning analytics tool that supports this strategy and worked with another group of teachers to define an ‘explainable learning analytics’ scheme that explains each cluster in a way that is both pedagogically meaningful and can be generated automatically. Third, we developed an AI algorithm that supports this ‘explainable clusters’ pedagogy and conducted a controlled experiment that evaluated its contribution to teachers’ ability to plan personalized learning sequences. The planned sequences were evaluated in a blinded fashion by an expert, and the results demonstrated that the experimental group – teachers who received the clusters with the explanations – designed sequences that addressed the difficulties exhibited by different groups of students better than those designed by teachers who received the clusters without explanations. The main contribution of this study is twofold. First, it presents an effective personalization approach that fits blended learning in the science classroom, which combines a real-time clustering algorithm with an explainable-AI scheme that can automatically build pedagogically meaningful explanations from item-level meta-data (Q Matrix). Second, it demonstrates how such an end-to-end learning analytics solution can be built with teachers through a co-design process and highlights the types of knowledge that teachers add to system-provided analytics in order to apply them to their local context. As a practical contribution, this process informed the design of a new learning analytics tool that was integrated into a free online learning platform that is being used by more than 1000 science teachers.

Evidence from various domains underlines the key role that human factors, and especially, trust, play in the adoption of technology by practitioners. In the case of Artificial Intelligence (AI) driven learning analytics tools, the issue is even more complex due to practitioners’ AI-specific misconceptions, myths, and fears (i.e., mass unemployment and ethical concerns). In recent years, artificial intelligence has been introduced increasingly into K-12 education. However, little research has been conducted on the trust and attitudes of K-12 teachers regarding the use and adoption of AI-based Educational Technology (EdTech).

AI, and, more specifically, deep learning, is profoundly impacting our industries and societies [1]. In recent years, machine learning (ML) 's surging impact has sparked discourse about the importance of AI education for young people, and in recent years, several initiatives and projects pursuing the mission of K-12 AI education have emerged. In 2020 Israel's Ministry of Education (MoE) approved a new comprehensive and rigorous ML curriculum targeting 11 and 12th-grade pupils majoring in computer science (CS). The curriculum is meant to be taught by the existing CS teacher workforce. However, since ML theory and practice are fundamentally different from traditional CS [2], implementing this thorough ML curriculum poses substantial challenges in developing an effective teaching workforce to deliver it. In this research, we suggest a solution for this challenge in the form of a blended-learning (BL) program. The online component of this program is based mainly on Coursera's Deep Learning Specialization MOOCs series [3]. The BL program, enhanced with pedagogical training, is also used for the professional development (PD) of the teachers who deliver the program. Out of fourteen CS teachers who participated in the PD in the summer of 2021, ten teach the BL program this year to 273 high-school pupils. Initial results demonstrate the achievement of the curriculum learning goals and provide compelling preliminary evidence that this program enables CS teachers who are new to machine learning to teach this thorough curriculum effectively

Learner-centered pedagogy highlights active learning and formative feedback. Instructors often incentivize learners to engage in such formative assessment activities by crediting their completion and score in the final grade, a pedagogical practice that is very relevant to MOOCs as well. However, previous studies have shown that too many MOOC learners exploit the anonymity to abuse the formative feedback, which is critical in the learning process, to earn points without effort. Unfortunately, limiting feedback and access to decrease cheating is counter-pedagogic and reduces the openness of MOOCs. We aimed to identify and analyze a MOOC assessment strategy that balances this tension between learner-centered pedagogy, incentive design, and reliability of the assessment. In this study, we evaluated an assessment model that MITx Biology introduced in a MOOC to reduce cheating with respect to its effect on two aspects of learner behavior - the amount of cheating and learners' engagement in formative course activities. The contribution of the paper is twofold. First, this work provides MOOC designers with an 'analytically-verified' MOOC assessment model to reduce cheating without compromising learner engagement in formative assessments. Second, this study provides a learning analytics methodology to approximate the effect of such an intervention.

The study presented in this paper deals with copying answers in MOOCs. Our findings show that a significant fraction of the certificate earners in the course that we studied have used what we call harvesting accounts to find correct answers that they later submitted in their main account, the account for which they earned a certificate. In total, around 2.5% of the users who earned a certificate in the course obtained the majority of their points by using this method, and around 10% of them used it to some extent. This paper has two main goals. The first is to define the phenomenon and demonstrate its severity. The second is characterizing key factors within the course that affect it, and suggesting possible remedies that are likely to decrease the amount of cheating. The immediate implication of this study is to MOOCs. However, we believe that the results generalize beyond MOOCs, since this strategy can be used in any learning environments that do not identify all registrants.

In this article, we discuss the possible connection between the programming language and the paradigm behind it, and programmers’ tendency to adopt an external or internal perspective of the system they develop. Based on a qualitative analysis, we found that when working with the visual, interobject language of live sequence charts (LSC), programmers tend to adopt an external and usability-oriented view of the system, whereas when working with an intraobject language, they tend to adopt an internal and implementation-oriented viewpoint. This is explained by first discussing the possible effect of the programming paradigm on programmers’ perception and then offering a more comprehensive explanation. The latter is based on a cognitive model of programming with LSC, which is an interpretation and a projection of the model suggested by Adelson and Soloway [1985] onto LSC and scenario-based programming, the new paradigm on which LSC is based. Our model suggests that LSC fosters a kind of programming that enables iterative refinement of the artifact with fewer entries into the solution domain. Thus, the programmer can make less context switching between the solution domain and the problem domain, and consequently spend more time in the latter. We believe that these findings are interesting mainly in two ways. First, they characterize an aspect of problem-solving behavior that to the best of our knowledge has not been studied before—the programmer’s perspective. The perspective can potentially affect the outcome of the problem-solving process, such as by leading the programmer to focus on different parts of the problem. Second, relating the structure of the language to the change in perspective sheds light on one of the ways in which the programming language can affect the programmer’s behavior.