Publications

As artificial intelligence (AI) becomes increasingly integrated into daily life, there is a critical need for developing AI literacy across all educational levels. However, current AI education remains largely confined to college-level computer science classrooms with limited access for K-12 learners. We present the AI Scholars Program, a novel approach that addresses the AI education gap by preparing college computing students to serve as AI education ambassadors in their communities and empowering K-12 teachers to adopt AI education practices in their classrooms. This experience report presents the curriculum and its outcomes after one round of refinement. The program offers structured AI learning through bi-weekly webinars, resources, and collaborative opportunities to form teams and conduct community outreach projects. Our program invited 63 scholars from 30 institutions across the U.S., including 51 college students and 12 K-12 teachers. Their outreach impacted over 230 K-12 learners. We examine program outcomes for participants and projects through pre/post surveys measuring computing attitudes and self-efficacy for teaching AI, scholar interviews, and outreach project reports. We share lessons learned and challenges for designing similar programs, highlighting the importance of involving educators for effective community-engaged AI education. The program creates a sustainable pipeline for college students to develop technical skills and leadership while addressing K-12 AI education shortages. We contribute insights for scaling AI literacy and broadening participation in computing.

The ICAP framework defines four cognitive engagement levels: Passive, Active, Constructive, and Interactive, where increased cognitive engagement can yield improved learning. However, personalizing learning activities that elicit the optimal level of cognitive engagement remains a key challenge in intelligent tutoring systems (ITS). In this work, we develop and evaluate a system that adaptively scaffolds cognitive engagement by dynamically selecting worked examples in two different ICAP modes:(active) Guided …

The ICAP framework defines four cognitive engagement levels: Passive, Active, Constructive, and Interactive, where increased cognitive engagement can yield improved learning. However, personalizing learning activities that elicit the optimal level of cognitive engagement remains a key challenge in intelligent tutoring systems (ITS). In this work, we develop and evaluate a system that adaptively scaffolds cognitive engagement by dynamically selecting worked examples in two different ICAP modes:(active) Guided …

As AI education expands in K-12 classrooms, there is a growing need to understand how AI learning experiences shape students’ motivational and affective orientations toward AI. Ability beliefs and intentions to persist are key constructs in AIED because they influence learners’ engagement in learning environments and continued participation. This study examines how a project-based AI learning experience affected these outcomes among middle school students. Ninety students completed a 10-hour AI module embedded within their science classes, in which they learned foundational AI concepts and designed conversational AI agents. Pre- and post-survey analyses showed increased ability beliefs but decreased intentions to persist in AI learning. This divergent pattern suggests a complex motivational response to hands-on AI instruction, indicating that gains in perceived competence do not necessarily translate into sustained persistence intentions. We discuss implications for the design of AI learning environments and for theoretical models of motivation and engagement in AI-supported learning contexts.

As Artificial Intelligence (AI) education has become a key component of K–12 curricula, activities such as designing and developing conversational agents are increasingly used as instructional practice. Prior work has primarily examined these activities by focusing on students’ learning outcomes or the quality of final AI artifacts, offering limited insight into the collaborative processes through which learning unfolds during AI system development. Although the AIED community has a long history of studying collaborative learning in STEM and Computing education, the emergence of AI learning environments in which students build AI systems presents new opportunities to understand how collaboration unfolds in AI education contexts. Grounded in these foundational works, the current study examines collaborative interaction among middle school students engaged in the design and development of an AI chatbot. Using Ordered Network Analysis of students’ dialogue and development actions, we characterize how collaboration is organized over time and how interaction patterns relate to chatbot quality and AI knowledge outcomes. Results reveal that higher-quality chatbots are associated with more integrated sequences linking explanation, testing, and refinement. Interaction patterns involving articulated reasoning and repeated testing and revision in response to chatbot output were also associated with stronger AI knowledge outcomes. These findings provide a process-oriented account of collaborative AI chatbot development and extend AIED research on collaborative learning processes to AI education contexts.

Effective tutoring requires distinguishing optimal, valid but suboptimal, and incorrect student solutions, a distinction central to intelligent tutoring systems (ITS) but untested for LLM-based tutors. As LLMs are increasingly explored as conversational complements to ITS, evaluating their diagnostic precision is essential. We present a benchmark of seven LLM feedback agents in propositional logic using knowledge-graph-derived ground truth across 10,836 solution–feedback pairs and three feedback condition s 1. Models achieved near-ceiling performance on optimal steps but systematically over-rejected valid but suboptimal reasoning and over-validated incorrect solutions, precisely where adaptive tutoring matters most. These failures persisted across models regardless of solution context, suggesting architectural rather than informational limits. Moreover, accurate diagnosis did not reliably produce pedagogically actionable feedback, revealing a gap between diagnostic judgment and instructional effectiveness. Our findings suggest that LLMs are better suited for hybrid architectures where KG-grounded models handle diagnosis while LLMs support open-ended scaffolding and dialogue.

This chapter explores the evolution of data-driven hint generation for intelligent tutoring systems (ITS). The Hint Factory and Interaction Networks have enabled the generation of next-step hints, waypoints, and strategic subgoals from historical student data. Data-driven techniques have also enabled systems to find the right time to provide hints. We explore further potential data-driven adaptations for problem solving based on behavioral problem solving data and the integration of Large Language Models (LLMs).

AI-based chatbots have the potential to accelerate learning and teaching, but may also have counterproductive consequences without thoughtful design and scaffolding. To better understand teachers’ perspectives on large language model (LLM) based chatbots, we conducted a study with 11 teams of middle-school teachers using chatbots for a science and computational thinking activity within a block-based programming environment. Based on a qualitative analysis of audio transcripts and chatbot interactions, we propose three profiles: explorer, frustrated, and mixed that reflect diverse scaffolding needs. In their discussions, we found that teachers perceived chatbot benefits such as building prompting skills and self confidence alongside risks including potential declines in learning and critical thinking. Key design recommendations include scaffolding the introduction to chatbots, facilitating teacher control of chatbot features, and suggesting when and how chatbots should be used. Our contribution informs the design of chatbots to support teachers and learners in middle school coding activities.

Tutoring systems improve learning through tailored interventions, such as worked examples, but often suffer from the aptitude-treatment interaction effect where low prior knowledge learners benefit more. We applied the ICAP learning theory to design two new types of worked examples, Buggy (students fix bugs), and Guided (students complete missing rules), requiring varying levels of cognitive engagement, and investigated their impact on learning in a controlled experiment with 155 undergraduate students in a logic problem solving tutor. Students in the Buggy and Guided examples groups performed significantly better on the posttest than those receiving passive worked examples. Buggy problems helped high prior knowledge learners whereas Guided problems helped low prior knowledge learners. Behavior analysis showed that Buggy produced more exploration-revision cycles, while Guided led to more help-seeking and fewer errors. This research contributes to the design of interventions in logic problem solving for varied levels of learner knowledge and a novel application of behavior analysis to compare learner interactions with the tutor.

Children increasingly interact with generative AI systems that can produce hallucinated content, potentially reinforcing misconceptions and undermining critical thinking skills. We investigate how children detect and respond to hallucinations while building and testing LLM-powered chatbots in a development environment. We integrated hallucination-awareness scaffolds such as confidence indicators, fact-checking, repeated questioning, and model comparison. Through a study with 48 middle school learners aged 10-14, participants showed significant pre-to-post gains in AI knowledge, hallucination awareness, and confidence in building trustworthy chatbots. They developed multi-layered strategies, including probing inconsistencies and cross-checking with external sources. Key challenges included over-reliance on visible cues, fragmented use of scaffolds, and a tension between creativity and reliability. These findings highlight design implications for children’s AI literacy for responsible AI development: supporting proactive, iterative engagement in the development cycle, integrating scaffolds into coherent workflows, and balancing creativity with accuracy.

Large language models (LLMs) are increasingly used for automated tutoring, but their reliability in structured symbolic domains remains unclear. We study step-level feedback for propositional logic proofs, which require precise symbolic reasoning aligned with a learner's current proof state. We introduce a knowledge-graph-grounded benchmark of 516 unique proof states with step-level annotations and difficulty metrics. Unlike prior tutoring evaluations that rely on model self-assessment or binary correctness, our framework enables fine-grained analysis of feedback quality against verified solution paths. We evaluate three role-specialized pipelines with varying solution access: Tutor (partial solution access), Teacher (full derivation access), and Judge (verification of Tutor feedback). Our results reveal a striking asymmetry: verification improves outcomes when upstream feedback is error-prone (<70% accuracy), but degrades performance by 4-6 percentage points through over-specification when feedback is already reliable (>85%). Critically, we identify a shared complexity ceiling; no model or pipeline reliably succeeds on proof states exceeding complexity 4-5. These findings challenge the assumption that adding verifiers or richer context universally improves tutoring, motivating adaptive, difficulty-aware architectures that route problems by estimated complexity and upstream reliability.

Project-based learning plays a crucial role in computing education. However, its open-ended nature makes tracking project development and assessing success challenging. We investigate how dialogue and system interaction logs predict project quality during collaborative, project-based AI learning of 94 middle school students working in pairs. We used linguistic features from dialogue transcripts and behavioral features from system logs to predict three project quality outcomes: productivity (number of training phrases), content richness (word density), and lexical variation (word diversity) of chatbot training phrases. We compared the predictive accuracy of each modality and a fusion of the modalities. Results indicate log data better predicts productivity, while dialogue data is more effective for content richness. Both modalities modestly predict lexical variation. Multimodal fusion improved predictions for productivity and lexical variation of training phrases but not content richness. These findings suggest that the value of multimodal fusion depends on the specific learning outcome. The study contributes to multimodal learning analytics by demonstrating the nuanced interplay between behavioral and linguistic data in assessing student learning progress in open-ended AI learning environments.

A long-standing challenge of intelligent tutoring systems (ITSs) is to determine when to provide what types of problems to learners based on their individual needs. In this study, we investigate a DRL-based adaptive pedagogical policy for choosing problem types among problem solving, worked examples, and Parsons’ problems in a data-driven intelligent logic tutor. We compare the DRL-based policy with a non-adaptive expert policy and a problem-solving-only control on learning gain and completion time. Overall, results show that the DRL policy did help adapt the tutor in some ways, significantly improving performance on one post-test problem and marginally reducing post-test time for low prior proficiency learners.

The variety of help that large language models (LLMs) provide has made them popular among students across fields. Computing education has been particularly affected, as LLMs can handle coding tasks effectively and provide feedback. This has raised hopes for supporting students, while creating concerns about learning and academic integrity. Researchers have responded by developing tools that leverage LLMs' potential while mitigating risks. Despite growing empirical studies on LLM-driven tools, there is no …

Linguistic alignment, the tendency of speakers to share common linguistic features during conversations, has emerged as a key area of research in computer-supported collaborative learning. While previous studies have shown that linguistic alignment can have a significant impact on collaborative outcomes, there is limited research exploring its role in K–12 learning contexts. This study investigates syntactic and lexical linguistic alignments in a collaborative computer science–learning corpus from 24 pairs (48 individuals) of middle school students (aged 11–13). The results show stronger effects of self-alignment than partner alignment on both syntactic and lexical levels, with students often diverging from their partners on task-relevant words. Furthermore, student self-alignment on the syntactic level is negatively correlated with partner satisfaction ratings, while self-alignment on lexical level is positively correlated with their partner’s satisfaction.

Confusion is a natural part of inquiry-based game-based learning environments (GBLE). While research has explored the relationships between single instances of confusion with motivation and learning, limited work has been done to understand overall confusion dynamics during long-running learning processes and their role in supporting students’ motivation. This study investigated the relationship between students’ confusion with end-of-game motivation and agency using gameplay and facial expression data collected from college students interacting with a GBLE for microbiology. The students were randomly assigned into two conditions; where in the Full Agency condition, students were allowed to move freely in the environment, whereas, in the Partial Agency condition, student actions were limited by a pedagogical scaffolding restricting possible actions to the ones considered most productive for learning. For students in the Full Agency condition (n=58), results showed an inverse relationship between the mean duration of confusion episodes during game-play and self-reported intrinsic motivation following game-play. However, these results were not observed for the Partial Agency condition (n=37), suggesting that pedagogical scaffolding that limits agency may also limit confusion in a GBLE. On the other hand, limiting agency may also limit opportunities to boost intrinsic motivation for students with well-developed self-regulated learning skills. These results suggest promising directions for research into the complex relationships between confusion, motivation, and agency.

Parsons problems (PPs) have shown promise in structured problem solving by providing scaffolding that decomposes the problem and requires learners to reconstruct the solution. However, some students face difficulties when first learning with PPs or solving more complex Parsons problems. This study introduces Guided Parsons problems (GPPs) designed to provide step-specific hints and improve learning outcomes in an intelligent logic tutor. In a controlled experiment with 76 participants, GPP students achieved significantly higher accuracy of rule application in both level-end tests and post-tests, with the strongest gains among students with lower prior knowledge. GPP students initially spent more time in training (1.52 vs. 0.81 hours) but required less time for post-tests, indicating improved problem solving efficiency. Our thematic analysis of GPP student self-explanations revealed task decomposition, better rule understanding, and reduced difficulty as key themes, while some students felt the structured nature of GPPs restricted their own way of reasoning. These findings reinforce that GPPs can effectively combine the benefits of worked examples and problem solving practice, but could be further improved by individual adaptation.

Block-Based Programming (BBP) platforms, such as Snap!, have become increasingly prominent in computer science education due to their ability to simplify programming concepts and foster computational thinking from an early age. While these platforms engage students through visual and gamified interfaces, teachers often face challenges in using them effectively and finding all the necessary features for classroom management. To address these challenges, we introduce SnapClass, a classroom management system integrated within the Snap! programming environment. SnapClass was iteratively developed drawing on established research about the pedagogical and logistical challenges teachers encounter in computing classrooms. Specifically, SnapClass allows educators to create and customize block-based coding assignments based on student skill levels, implement rubric-based auto-grading, and access student code history and recovery features. It also supports monitoring student engagement and idle time, and includes a help dashboard with a “raise hand” feature to assist students in real time. This paper describes the design and key features of SnapClass those are developed and those are under progress.

Intelligent tutoring systems have demonstrated effectiveness in teaching formal propositional logic proofs, but their reliance on template-based explanations limits their ability to provide personalized student feedback. While large language models (LLMs) offer promising capabilities for dynamic feedback generation, they risk producing hallucinations or pedagogically unsound explanations. We evaluated the stepwise accuracy of LLMs in constructing multi-step symbolic logic proofs, comparing six prompting techniques across …

Developers rarely build programming environments that help secondary teachers support student learning. We interviewed 11 K12 teachers to discover how they support students learning to program and how tools might assist their teaching practice. Based on thematic analysis and organizing teacher activities around student actions, we have derived a new framework that can be used to design a programming learning system to support teachers. Our results suggest that teachers structure their activities based on their ideals about effective programming teaching and learning, and student problem solving and help-seeking processes. Therefore, our framework relates the themes we discovered about teacher activities to ideals and student problem solving in a time-based framework that can inform the design for new programming learning systems.

As artificial intelligence (AI) becomes more prominent in children's lives, an increasing number of researchers and practitioners have underscored the importance of integrating AI as learning content in K-12. Despite the recent efforts in developing AI curricula and guiding frameworks in AI education, the educational opportunities often do not provide equally engaging and inclusive learning experiences for all learners. To promote equality and equity in society and increase competitiveness in the AI workforce, it is essential to broaden participation in AI education. However, a framework that guides teachers and learning designers in designing inclusive learning opportunities tailored for AI education is lacking. Universal Design for Learning (UDL) provides guidelines for making learning more inclusive across disciplines. Based on the principles of UDL, this paper proposes a framework to guide the design of inclusive AI learning. We conducted a systematic literature review to identify AI learning design-related frameworks and synthesized them into our proposed framework, which includes the core component of AI learning content (i.e., five big ideas), anchored by the three UDL principles (the “why,” “what,” and “how” of learning), and six praxes with pedagogical examples of AI instruction. Alongside this, we present an illustrative example of the application of our proposed framework in the context of a middle school AI summer camp. We hope this paper will guide researchers and practitioners in designing more inclusive AI learning experiences.

As conversational AI apps such as Siri and Alexa become ubiquitous among children, the CS education community has begun leveraging this popularity as a potential opportunity to attract young learners to AI, CS, and STEM learning. However, teaching conversational AI to K-12 learners remains challenging and unexplored due in part to the abstract and complex nature of some conversational AI concepts, such as intents and training phrases. One promising approach to teaching complex topics in engaging ways is through unplugged activities, which have been shown to be highly effective in fostering CS conceptual understanding without using computers. Research efforts are underway toward developing unplugged activities for teaching AI, but few thus far have focused on conversational AI. This experience report describes the design and iterative refinement of a series of novel unplugged activities for a conversational AI summer camp for middle school learners. We discuss learner responses and lessons learned through our implementation of these unplugged activities. Our hope is that these insights support CS education researchers in making conversational AI learning more engaging and accessible to all learners.

Task-oriented dialogue systems are expected to handle a constantly expanding set of intents and domains even after they have been deployed to support more and more functionalities. To live up to this expectation, it becomes critical to mitigate the catastrophic forgetting problem (CF) that occurs in continual learning (CL) settings for a task such as intent recognition. While existing dialogue systems research has explored replay-based and regularization-based methods to this end, the effect of domain ordering on the CL performance of intent recognition models remains unexplored. If understood well, domain ordering has the potential to be an orthogonal technique that can be leveraged alongside existing techniques such as experience replay. Our work fills this gap by comparing the impact of three domain-ordering strategies (min-sum path, max-sum path, random) on the CL performance of a generative intent recognition model. Our findings reveal that the min-sum path strategy outperforms the others in reducing catastrophic forgetting when training on the 220M T5-Base model. However, this advantage diminishes with the larger 770M T5-Large model. These results underscores the potential of domain ordering as a complementary strategy for mitigating catastrophic forgetting in continually learning intent recognition models, particularly in resource-constrained scenarios.

Recent advancements in automatic evaluation have made significant progress, yet evaluating learner-created computational artifacts such as project-based code remains challenging. This study investigates the capability of GPT-4, a state-of-the-art Large Language Model (LLM), in assessing learner-created computational artifacts. Specifically, we analyze the source code of 75 chatbots predominantly built by middle school learners. We compare four LLM prompting strategies ranging from example-based to rubric-informed approaches. The experimental results indicate that the LLM-based evaluation module achieves substantial agreement (Cohen’s weighted= 0.797) with human evaluators in two of five artifact dimensions, moderate agreement in one, and fair agreement in the remaining two dimensions. We analyze the trade-offs between different LLM prompting strategies through qualitative error analysis. The findings demonstrate the potential of LLMs for automatically evaluating project-based, open-ended computational artifacts.

Previous literature has associated math literacy with linguistic factors such as verbal ability and phonological skills. However, few studies have investigated linguistic synchrony, shown in mathematical discussions. This study modelled math literacy and examined the relationship of math literacy with linguistic synchrony between students and facilitators. We retrieved data from 20,776 online mathematical discussion threads at a secondary school level. First, we assessed students' math literacy based on their discussions and classified them into high‐ and low‐math literacy groups. Then, we conducted Cross‐Recurrence Quantification Analysis (CRQA) to calculate linguistic synchrony within each thread. The result implies that students with high math literacy are more likely to share common words (eg, mathematical terms) with facilitators. At the same time, they would paraphrase the facilitators' words rather than blindly mimic them as the exact sentences or phrases. On the other hand, students with low math literacy tend to use overlapping words with facilitators less frequently and are more likely to repeat the exact same phrases from the facilitators. The findings provide an empirical data analysis and insights into mathematical discussions and linguistic synchrony. In addition, this paper implies the directions to improve online mathematical discussions and foster math literacy. Practitioner notes What is already known about this topic Mathematical discussions are known to be an effective way to promote math literacy. Math literacy and linguistic skills have a strong link. Linguistic synchrony is related to better collaboration and common knowledge building. What this paper adds Reveals the relationship between math literacy and linguistic synchrony and deepens the understanding of digital communication in online learning environments. Provides empirical analysis of natural language data in group discussions using CRQA. Conceptualizes linguistic synchrony with three sub‐concepts: linguistic concurrence, predictability, and complexity. Implications for practice and/or policy Educators and practitioners could utilize the automatic formative assessment of math literacy based on the student's language use in mathematical discussions. Educational technology researchers and designers could include CRQA indices and recurrence plots in the dashboard design to provide information to support teachers and learners. Teachers would be able to provide real‐time interventions to promote effective mathematical communication and foster math literacy throughout mathematical discussions. What is already known about this topic Mathematical discussions are known to be an effective way to promote math literacy. Math literacy and linguistic skills have a strong link. Linguistic synchrony is related to better collaboration and common knowledge building. What this paper adds Reveals the relationship between math literacy and linguistic synchrony and deepens the understanding of digital communication in online learning environments. Provides empirical analysis of natural language data in group discussions using CRQA. Conceptualizes linguistic synchrony with three sub‐concepts: linguistic concurrence, predictability, and complexity. Implications for practice and/or policy Educators and practitioners could utilize the automatic formative assessment of math literacy based on the student's language use in mathematical discussions. Educational technology researchers and designers could include CRQA indices and recurrence plots in the dashboard design to provide information to support teachers and learners. Teachers would be able to provide real‐time interventions to promote effective mathematical communication and foster math literacy throughout mathematical discussions.

Natural Language Processing (NLP) plays a significant role in our daily lives and has become an essential part of Artificial Intelligence (AI) education in K-12. As children grow up with NLP-powered applications, it is crucial to introduce NLP concepts to them, fostering their understanding of language processing, language generation, and ethical implications of AI and NLP. This paper presents a comprehensive review of digital learning environments for teaching NLP in K-12. Specifically, it explores existing digital learning tools, discusses how they support specific NLP tasks and procedures, and investigates their explainability and evaluation results in educational contexts. By examining the strengths and limitations of these tools, this literature review sheds light on the current state of NLP learning tools in K-12 education. It aims to guide future research efforts to refine existing tools, develop new ones, and explore more effective and inclusive strategies for integrating NLP into K-12 educational contexts.

The ubiquity of AI-based conversational apps such as Siri, Alexa and Google Assistant means more young users are interacting with these apps. The increasing popularity of these conversational applications brings a potential opportunity to attract learners to AI, CS and STEM fields. CS Education researchers need to explore how to leverage this opportunity, in particular to serve learners who are underrepresented in CS and STEM. This experience report describes the design and iterative refinement of a series of two-week summer camps in which 62 predominantly Black students participated in hands-on AI-based learning experiences to design and develop their own conversational AI apps. We discuss the organization of this summer camp experience, including strategies for recruiting from and building trust within the target community, designing professional development for camp facilitators, structuring the camp activities, and encouraging projects that are personally and socially relevant. We share challenges and lessons learned from this AI summer camp in the hopes that they will inform other researchers and practitioners who are interested in designing and deploying similar experiences.

As artificial intelligence permeates our lives through various tools and services, there is an increasing need to consider how to teach young learners about AI in a relevant and engaging way. One way to do so is to leverage familiar and pervasive technologies such as conversational AIs. By learning about conversational AIs, learners are introduced to AI concepts such as computers’ perception of natural language, the need for training datasets, and the design of AI-human interactions. In this experience report, we describe a summer camp curriculum designed for middle school learners composed of general AI lessons, unplugged activities, conversational AI lessons, and project activities in which the campers develop their own conversational agents. The results show that this summer camp experience fostered significant increases in learners’ ability beliefs, willingness to share their learning experience, and intent to persist in AI learning. We conclude with a discussion of how conversational AI can be used as an entry point to K-12 AI education.

Abstract Conversational AIs such as Alexa and ChatGPT are increasingly ubiquitous in young people's lives, but these young users are often not afforded the opportunity to learn about the inner workings of these technologies. One of the most powerful ways to foster this learning is to empower youth to create AI that is personally and socially meaningful to them. We have built a novel development environment, AMBY–“AI Made By You”–for youth to create conversational agents. AMBY was iteratively designed with and for youth aged 12–13 …

Mathematical discussions have become a popular educational strategy to promote math literacy. While some studies have associated math literacy with linguistic factors such as verbal ability and phonological skills, no studies have examined the relationship between linguistic synchrony and math literacy. In this study, we modeled linguistic synchrony and students’ math literacy from 20,776 online mathematical discussion threads between students and facilitators. We conducted Cross-Recurrence Quantification Analysis (CRQA) to calculate linguistic synchrony within each thread. The statistical testing result comparing CRQA indices between high and low math literacy groups shows that students with high math literacy have a significantly higher Recurrence Rate (RR), Number of Recurrence Lines (NRLINE), and the average Length of lines (L), but lower Determinism (DET) and normalized Entropy (rENTR). This result implies that students with high math literacy are more likely to share common words with facilitators, but they would paraphrase them. On the other hand, students with low math literacy tend to repeat the exact same phrases from the facilitators. The findings provide a better understanding of mathematical discussions and can potentially guide teachers in promoting effective mathematical discussions.

Summer camps have become popular for introducing K-12 learners to computer science (CS) and artificial intelligence (AI) in informal learning environments. Facilitators play crucial roles in guiding and engaging learners in these contexts, but there is limited research on their roles in informal AI learning. This paper examines facilitators’ dialogues with campers in a middle school AI summer camp, identifying eight major facilitator roles. The roles differed depending on group dynamics and project phase. The paper provides empirical grounding to define facilitators’ roles in AI learning and guide the design of professional development for camp facilitators.

Power dynamics in human-human communication can impact rapport-building and learning gains, but little is known about how power impacts human-agent communication. In this paper, we examine dominance behavior in utterances between middle-school students and a teachable robot as they work through math problems, as coded by Rogers and Farace's Relational Communication Control Coding Scheme (RCCCS). We hypothesize that relatively dominant students will show increased learning gains, as will students with greater dominance agreement with the robot. We also hypothesize that gender could be an indicator of difference in dominance behavior. We present a preliminary analysis of dominance characteristics in some of the transactions between robot and student. Ultimately, we hope to determine if manipulating the dominance behavior of a learning robot could support learning.

More young people are interacting with smart conversational agents such as Alexa and Google Assistant. These platforms are extensible, providing, in principle, a compelling opportunity for young users to create and tinker with their own conversational agents. However, to date the interfaces for conversational app development are adult-focused. This paper presents the early design process for AMBY (AI Made by You), which we are building to empower young learners to create their own conversational agents. We first conducted a contextual inquiry with 14 middle school students (aged 11-13) in an AI summer camp, followed by two other usability studies. The system design has been refined after each study. Key features of AMBY include a visual dialogue management panel, testing panel with a diverse avatar, and a voice input modality. AMBY is designed to serve as a pedagogically-robust resource for K-12 AI education and as an engaging and creative way for middle schoolers to explore AI.

This study examined undergraduate STEM students’ experiences using an online introductory computer programming learning environment equipped with an automated hint generation system. Following a convergent parallel mixed methods design, this study utilized both quantitative and qualitative data from student experiential data. Analysis by level of prior knowledge demonstrated that elements of the learning environment did not cater to their learning needs and cognitive architecture. Cognitive Load Theory was used to contextualize system elements against both higher and lower prior experience learners, ultimately pointing to a need to design better scaffolds and hints to the needs of novice CS learners.

Research has shown study habits and skills to be correlated with academic success, calling for a deeper comprehension of these behaviors and processes to design effective interventions for struggling students. Chatbots have recently been used as a persuasive technology to help support behavioral change, making them an intriguing design space for students' study habits and skills. This paper investigated the feasibility of using chatbots for promoting behavioral change of college students majoring in Computer Science (CS). We conducted semi-structured interviews with CS peer-tutors and surveyed university freshmen to understand students' study habits and identify technical intervention opportunities. Inspired by the findings, we designed StudyBuddy, a chatbot prototype deployed in Slack that periodically sends tips, provides assessments of students' study habits via surveys, helps the students break down assignments, recommends academic resources, and sends reminders. We evaluated the usability of the prototype in-depth with 8 students (both first-year and senior students) and 5 course instructors followed by a large scale evaluative survey (n=117) using video of the prototype. Our research identified important design challenges such as building trust and preserving privacy, limiting interaction costs, and supporting both immediate and long-term sustainable support. Likewise, we proposed design recommendations that demonstrate context awareness, personalize the experience based on user preferences, and adapt over time as students mature and grow.

Modeling a learner’s frustration in adaptive environments can inform scaffolding. While much work has explored momentary frustration, there is limited research investigating the dynamics of frustration over time and its relationship with problem-solving behaviors. In this paper, we clustered 86 undergraduate students into four frustration trajectories as they worked with an adaptive learning environment for introductory computer science. The results indicate that students who initially report high levels of frustration but then reported lower levels later in their problem solving were more likely to have sought help. These findings provide insight into how frustration trajectory models can guide adaptivity during extended problem-solving episodes.

Block-based programming environments are widely used in computer science education. However, these environments pose significant challenges for student modeling. Given a series of problem-solving actions taken by students in block-based programming environments, student models need to accurately infer problem-solving students’ programming abilities in real time to enable adaptive feedback and hints that are tailored to students’ abilities. While student models for block-based programming offer the potential to support student-adaptivity, creating student models for these environments is challenging because students can develop a broad range of solutions to a given programming activity. To address these challenges, we introduce a progression trajectory-based student modeling framework for modeling novice student block-based programming across multiple learning activities. Student trajectories utilize a time series representation that employs code analysis to incrementally compare student programs to expert solutions as students undertake block-based programming activities. This paper reports on a study in which progression trajectories were collected from more than 100 undergraduate students engaging in a series of block-based programming activities in an introductory computer science course. Using progression trajectory-based student modeling, we identified three distinct trajectory classes: Early Quitting, High Persistence, and Efficient Completion. Analysis of these trajectories revealed that they exhibit significantly different characteristics with respect to students’ actions and can be used to accurately predict students’ programming behaviors on future programming activities compared to competing baseline models. The findings suggest that progression trajectory-based student models can accurately model students’ block-based programming problem solving and hold potential for informing adaptive support in block-based programming environments.

Social robots have been shown to be effective educational tools. Rapport, or interpersonal closeness, can lead to better human-robot interactions and positive learning outcomes. Prior research has investigated the effects of social robots on student rapport and learning in a single session, but little is known about how individuals build rapport with a robot over multiple sessions. We reported on a case study in which 7 middle school students explained mathematics concepts to an intelligent teachable robot named Emma for five sessions. We modeled learners’ rapport-building linguistic strategies to understand whether the ways middle school students build rapport with the robot over time follow the same trends as human conversation, and how individual differences might mediate the rapport between human and robot.

Tangible programming lets children build programs with physical blocks, yet many still struggle to debug and trace how their sequences execute. AI has the potential to provide personalized support, but we know little about when children want help or what types of help they find useful. We conducted an exploratory study with 16 children using a tangible programming prototype with an embedded AI assistant, situated in augmented reality (AR). Our results show that children treated AI as a situational resource, using it mainly when they were stuck or seeking new ideas. They also faced challenges with the AI's contextual understanding and communication, including breakdowns that stemmed from AR sensing. Children envisioned AI that is embodied in the AR scene, grounds help in the tangible workspace (eg, pointing to specific blocks or execution steps), and takes actionable in-system steps (eg, creating virtual copies or placeholders for physical blocks). We discuss the implications for child-centered AI support in tangible programming and reflect on how children's use and non-use of AI express their agency in AR-based physical interfaces.