TY - JOUR TI - Tackling the Credit Assignment Problem in Reinforcement Learning-Induced Pedagogical Policies with Neural Networks AU - Ausin, Markel Sanz AU - Maniktala, Mehak AU - Barnes, Tiffany AU - Chi, Min T2 - ARTIFICIAL INTELLIGENCE IN EDUCATION (AIED 2021), PT I AB - Intelligent Tutoring Systems (ITS) provide a powerful tool for students to learn in an adaptive, personalized, and goal-oriented manner. In recent years, Reinforcement Learning (RL) has shown to be capable of leveraging previous student data to induce effective pedagogical policies for future students. One of the most desirable goals of these policies is to maximize student learning gains while minimizing the training time. However, this metric is often not available until a student has completed the entire tutor. For this reason, the reinforcement signal of the effectiveness of the tutor is delayed. Assigning credit for each intermediate action based on a delayed reward is a challenging problem denoted the temporal Credit Assignment Problem (CAP). The CAP makes it difficult for most RL algorithms to assign credit to each action. In this work, we develop a general Neural Network-based algorithm that tackles the CAP by inferring immediate rewards from delayed rewards. We perform two empirical classroom studies, and the results show that this algorithm, in combination with a Deep RL agent, can improve student learning performance while reducing training time. DA - 2021/// PY - 2021/// DO - 10.1007/978-3-030-78292-4_29 VL - 12748 SP - 356-368 SN - 1611-3349 KW - Pedagogical agent KW - Credit assignment problem KW - Deep reinforcement learning ER - TY - JOUR TI - Evaluating Critical Reinforcement Learning Framework in the Field AU - Ju, Song AU - Zhou, Guojing AU - Abdelshiheed, Mark AU - Barnes, Tiffany AU - Chi, Min T2 - ARTIFICIAL INTELLIGENCE IN EDUCATION (AIED 2021), PT I AB - Reinforcement Learning (RL) is learning what action to take next by mapping situations to actions so as to maximize cumulative rewards. In recent years RL has achieved great success in inducing effective pedagogical policies for various interactive e-learning environments. However, it is often prohibitive to identify the critical pedagogical decisions that actually contribute to desirable learning outcomes. In this work, by utilizing the RL framework we defined critical decisions to be those states in which the agent has to take the optimal actions, and subsequently, the Critical policy as carrying out optimal actions in the critical states while acting randomly in others. We proposed a general Critical-RL framework for identifying critical decisions and inducing a Critical policy. The effectiveness of our Critical-RL framework is empirically evaluated from two perspectives: whether optimal actions must be carried out in critical states (the necessary hypothesis) and whether only carrying out optimal actions in critical states is as effective as a fully-executed RL policy (the sufficient hypothesis). Our results confirmed both hypotheses. DA - 2021/// PY - 2021/// DO - 10.1007/978-3-030-78292-4_18 VL - 12748 SP - 215-227 SN - 1611-3349 KW - Critical decisions KW - Reinforcement learning KW - ITS ER - TY - JOUR TI - Infusing Computing: Moving a Service Oriented Internship Program Online AU - Isvik, Amy AU - Catete, Veronica AU - Bell, Dave AU - Gransbury, Isabella AU - Barnes, Tiffany T2 - IEEE STCBP RESPECT CONFERENCE: 2021 RESEARCH ON EQUITY AND SUSTAINED PARTICIPATION IN ENGINEERING, COMPUTING, AND TECHNOLOGY (RESPECT) AB - As virtual conferencing technology becomes more common and situations make in-person experiences difficult or unsafe to host, the need for online internships to support sustained participation in computing increases. We investigate the problem of how to provide a meaningful experiential education program in a virtual environment and serve geographically dispersed participants through our experience with moving a service oriented internship program online. Our computer science internship program leverages high school interns' programming skills and classroom experience to assist teachers in developing computing-infused lessons for their classrooms. Using a combination of synchronous and asynchronous activities, we trained our interns in how to make these lessons and helped interns build community amongst themselves. Our interns created over 90 lessons during the summer and helped over 50 teachers create their own lessons at an infusing computing professional development. DA - 2021/// PY - 2021/// DO - 10.1109/RESPECT51740.2021.9620644 SP - 199-203 KW - computing education KW - virtual internship KW - service-learning ER - TY - JOUR TI - Examining Equity in Computing-Infused Lessons Made by Novices AU - Isvik, Amy AU - Catete, Veronica AU - Elmore, Erynn AU - Barnes, Tiffany T2 - IEEE STCBP RESPECT CONFERENCE: 2021 RESEARCH ON EQUITY AND SUSTAINED PARTICIPATION IN ENGINEERING, COMPUTING, AND TECHNOLOGY (RESPECT) AB - In this study, we examine 10 computing-infused lessons with high equity scores created by high school interns. These projects were part of a larger corpus of 90+ projects made in summer 2020 for middle school and high school classrooms and the projects were evaluated using the Teacher Accessibility, Equity, and Content (TEC) rubric. This article examines the observed extensive evidence for equity in these 10 projects to determine how meaningful these equity scores are, what themes are present across projects, and to provide curriculum developers with strategies for ensuring their activities utilize equitable practices to be intentionally inclusive of all students. DA - 2021/// PY - 2021/// DO - 10.1109/RESPECT51740.2021.9620700 SP - 157-161 KW - equity KW - curriculum design KW - equity analysis ER - TY - JOUR TI - To Reduce Healthcare Workload: Identify Critical Sepsis Progression Moments through Deep Reinforcement Learning AU - Ju, Song AU - Kim, Yeo Jin AU - Ausin, Markel Sanz AU - Mayorga, Maria E. AU - Chi, Min T2 - 2021 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA) AB - Healthcare systems are struggling with increasing workloads that adversely affect quality of care and patient outcomes. When clinical practitioners have to make countless medical decisions, they may not always able to make them consistently or spend time on them. In this work, we formulate clinical decision making as a reinforcement learning (RL) problem and propose a human-controlled machine-assisted (HC-MA) decision making framework whereby we can simultaneously give clinical practitioners (the humans) control over the decision-making process while supporting effective decision-making. In our HC-MA framework, the role of the RL agent is to nudge clinicians only if they make suboptimal decisions at critical moments. This framework is supported by a general Critical Deep RL (Critical-DRL) approach, which uses Long-Short Term Rewards (LSTRs) and Critical Deep Q-learning Networks (CriQNs). Critical-DRL’s effectiveness has been evaluated in both a GridWorld game and real-world datasets from two medical systems: a large health system in the northeast of USA, referred as NEMed and Mayo Clinic in Rochester, Minnesota, USA for septic patient treatment. We found that our Critical-DRL approach, by which decisions are made at critical junctures, is as effective as a fully executed DRL policy and moreover, it enables us to identify the critical moments in the septic treatment process, thus greatly reducing burden on medical decision-makers by allowing them to make critical clinical decisions without negatively impacting outcomes. DA - 2021/// PY - 2021/// DO - 10.1109/BigData52589.2021.9671407 SP - 1640-1646 SN - 2639-1589 KW - Reinforcement Learning KW - Sepsis KW - Critical Decision ER - TY - JOUR TI - InferNet for Delayed Reinforcement Tasks: Addressing the Temporal Credit Assignment Problem AU - Ausin, Markel Sanz AU - Azizsoltani, Hamoon AU - Ju, Song AU - Kim, Yeo Jin AU - Chi, Min T2 - 2021 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA) AB - Rewards are the critical signals for Reinforcement Learning (RL) algorithms to learn the desired behavior in a sequential multi-step learning task. However, when these rewards are delayed and noisy in nature, the learning process becomes more challenging. The temporal Credit Assignment Problem (CAP) is a well-known and challenging task in AI. While RL, especially Deep RL, often works well with immediate rewards but may fail when rewards are delayed or noisy, or both. In this work, we propose delegating the CAP to a Neural Network-based algorithm named InferNet that explicitly learns to infer the immediate rewards from the delayed and noisy rewards. The effectiveness of InferNet was evaluated on three online RL tasks: a GridWorld, a CartPole, and 40 Atari games; and two offline RL tasks: GridWorld and a real-life Sepsis treatment task. The effectiveness of InferNet rewards is compared to that of immediate and delayed rewards in two settings: with and without noise. For the offline RL tasks, it is also compared to a strong baseline, InferGP [7]. Overall, our results show that InferNet is robust to delayed or noisy reward functions, and it could be used effectively for solving the temporal CAP in a wide range of RL tasks, when immediate rewards are not available or they are noisy. DA - 2021/// PY - 2021/// DO - 10.1109/BigData52589.2021.9671827 SP - 1337-1348 SN - 2639-1589 KW - Credit Assignment Problem KW - Deep Reinforcement Learning ER - TY - JOUR TI - Multi-Temporal Abstraction with Time-Aware Deep Q-Learning for Septic Shock Prevention AU - Kim, Yeo Jin AU - Ausin, Markel Sanz AU - Chi, Min T2 - 2021 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA) AB - Sepsis is a life-threatening organ dysfunction and a disease of astronomical burden. Septic shock, the most severe complication of sepsis, leads to a mortality rate as high as 50%. However, septic shock prevention is extremely challenging because individual patients often have very different disease progression, and thus the timings of medical interventions can play a key role in their effectiveness. Recently, reinforcement learning (RL) methods like deep Q-learning networks (DQN) have shown great promise in developing effective treatments for preventing septic shock. In this work, we propose MTA-TQN, a Multi-view -Temporal Abstraction mechanism within a Time-aware deep Q-learning Network framework for this task. More specifically, 1) MTA-TQN leverages irregular time intervals to discount expected return which would prevent systemic overestimations caused by temporal discount errors; 2) it learns both short and long-range dependencies with multi-view temporal abstractions which would reduce bias to a specific series of observations for a single state. The effectiveness of MTA-TQN is validated on two hard exploration Atari games and the septic shock prevention task using real-world EHRs. Our results demonstrate that both time-awareness and multi-view temporal abstraction are essential to induce effective policies, particularly with irregular time-series data. In the septic shock prevention task, while the top 10% of patients whose treatments agreed with DQN induced policy experienced a 17% septic shock rate, our MTA-TQN policies achieved a 5.7% septic shock rate. DA - 2021/// PY - 2021/// DO - 10.1109/BigData52589.2021.9671662 SP - 1657-1663 SN - 2639-1589 KW - deep reinforcement learning KW - time-aware KW - temporal abstraction KW - sepsis ER - TY - JOUR TI - Removing the Walls Around Visual Educational Programming Environments AU - Broll, Brian AU - Ledeczi, Akos AU - Stein, Gordon AU - Jean, Devin AU - Brady, Corey AU - Grover, Shuchi AU - Catete, Veronica AU - Barnes, Tiffany T2 - 2021 IEEE SYMPOSIUM ON VISUAL LANGUAGES AND HUMAN-CENTRIC COMPUTING (VL/HCC 2021) AB - Many block-based programming environments have proven to be effective at engaging novices in learning programming. However, most restrict access to the outside world, limiting learners to commands and computing resources built in to the environment. Some allow learners to drag and drop files, connect to sensors and robots locally or issue HTTP requests. But in a world where most of the applications in our daily lives are distributed (i.e., their functionality depends on communicating with other programs or accessing resources and data on the internet), the lack of support for beginners to envision and create such distributed programs is a lost opportunity. This paper argues that it is not only feasible, but crucial, to create environments with simple yet powerful abstractions that open up distributed computing and other widely used but advanced computing concepts including networking, the Internet of Things, and cybersecurity to novices. By thus removing the walls around our environments, we can expand opportunities for learning considerably: programs can access a wealth of online data and web services, and communicate with other projects. Moreover, these changes can enable young learners to collaborate with each other during program construction whether they share their physical location or study remotely. Importantly, providing access to the wider world will also help counter widespread student perceptions that block-based environments are mere toys, and show that they are capable of creating compelling applications. The paper presents NetsBlox, a programming environment that supports these ideas and shows that tools can be designed to democratize access to powerful ideas in computing. DA - 2021/// PY - 2021/// DO - 10.1109/VL/HCC51201.2021.9576399 SP - SN - 1943-6092 ER - TY - JOUR TI - PEDI - Piazza Explorer Dashboard for Intervention AU - Akintunde, Ruth Okoilu AU - Limke, Ally AU - Barnes, Tiffany AU - Heckman, Sarah AU - Lynch, Collin T2 - 2021 IEEE SYMPOSIUM ON VISUAL LANGUAGES AND HUMAN-CENTRIC COMPUTING (VL/HCC 2021) AB - Analytics about how students navigate online learning tools throughout the duration of an assignment is scarce. Knowledge about how students use online tools before a course's end could positively impact students' learning outcomes. We introduce PEDI (Piazza Explorer Dashboard for Intervention), a tool which analyzes and presents visualizations of forum activity on Piazza, a question and answer forum, to instructors. We outline the design principles and data-informed recommendations used to design PEDI. Our prior research revealed two critical periods in students' forum engagement over the duration of an assignment. Early engagement in the first half of an assignment duration positively correlates with class average performance. Whereas, extremely high engagement toward the deadline predicted lower class average performance. PEDI uses these findings to detect and flag troubling engagement levels and informs instructors through clear visualizations to promote data-informed interventions. By providing insights to instructors, PEDI may improve class performance and pave the way for a new generation of online tools. DA - 2021/// PY - 2021/// DO - 10.1109/VL/HCC51201.2021.9576443 SP - SN - 1943-6092 KW - learning analytics dashboards KW - forum activity KW - real time visualizations ER - TY - JOUR TI - You Really Need Help: Exploring Expert Reasons for Intervention During Block-based Programming Assignments AU - Dong, Yihuan AU - Shabrina, Preya AU - Marwan, Samiha AU - Barnes, Tiffany T2 - ICER 2021: PROCEEDINGS OF THE 17TH ACM CONFERENCE ON INTERNATIONAL COMPUTING EDUCATION RESEARCH AB - In recent years, research has increasingly focused on developing intelligent tutoring systems that provide data-driven support for students in need of assistance during programming assignments. One goal of such intelligent tutors is to provide students with quality interventions comparable to those human tutors would give. While most studies focused on generating different forms of on-demand support, such as next-step hints and worked examples, at any given moment during the programming assignment, there is a lack of research on why human tutors would provide different forms of proactive interventions to students in different situations. This information is critical to know to allow the intelligent programming environments to select the appropriate type of student support at the right moment. DA - 2021/// PY - 2021/// DO - 10.1145/3446871.3469764 SP - 334-346 KW - novice programming KW - proactive intervention KW - block-based environments KW - programming assignments KW - expert intervention ER - TY - JOUR TI - Exploring and Influencing Teacher Grading for Block-based Programs through Rubrics and the GradeSnap Tool AU - Milliken, Alexandra AU - Catete, Veronica AU - Limke, Ally AU - Gransbury, Isabella AU - Chipman, Hannah AU - Dong, Yihuan AU - Barnes, Tiffany T2 - ICER 2021: PROCEEDINGS OF THE 17TH ACM CONFERENCE ON INTERNATIONAL COMPUTING EDUCATION RESEARCH AB - This article examines the grading process and profiles of secondary computer science teachers as they assess block-based student programming submissions. Through an iterative design process, we have created a new tool, Gradesnap, which streamlines how teachers can open, review, and evaluate student submissions within the same interface. Our study compares teachers’ grading processes using the different assessment formats, so that we can understand how their grading processes can be augmented or supported to reduce ’pain points’ and to enable teachers to provide more constructive and formative feedback for students. We use a case study approach to examine the experiences and outcomes of four secondary computer science teachers with varied teaching and assessment experience, when grading as usual, grading with a rubric, and grading with GradeSnap. Our study shows that when participants use GradeSnap, they are able to give supportive comments to lower performing and borderline students who need critical feedback to better understand misconceptions. We also discovered that the different grading processes provided a vehicle for reflection for some teachers in understanding their grading goals and how they enact them. This research is the first to examine teacher grading processes for computer science, and highlights the need for teacher preparation and support for providing programming feedback and assessment. DA - 2021/// PY - 2021/// DO - 10.1145/3446871.3469762 SP - 101-114 KW - block-based languages KW - grading and assessment tools KW - secondary teacher tools ER - TY - JOUR TI - Data to Donations: Towards In-Kind Food Donation Prediction across Two Coasts AU - Sharma, Esha AU - Davis, Lauren AU - Ivy, Julie AU - Chi, Min T2 - 2021 IEEE GLOBAL HUMANITARIAN TECHNOLOGY CONFERENCE (GHTC) AB - Our goal in this work is to build effective yet robust models to predict unreliable and inconsistent in-kind donations at both weekly and monthly levels for two food banks across coasts: the Food Bank of Central Eastern North Carolina in North Carolina and Los Angeles Regional Food Bank in California. We explore three factors: model, data length, and window type. For the model, we evaluate a series of classic time-series forecasting models against the state-of-the-art approaches such as Bayesian Structural Time Series modeling (BSTS) and deep learning models; for the data length, we vary training data from 2 weeks to 13 years; for the window type, we compare sliding vs. expanding. Our results show the effectiveness of different models heavily depends on the data length and the window type as well as characteristics of the food bank. Motivated by these findings, we investigate the effectiveness of employing an average of all predictions formed by considering all three factors at both monthly and weekly levels for both food banks. Our results show that this average of predictions significantly and consistently outperforms all classical models, deep learning, and BSTS for the donation prediction at both monthly and weekly levels for both food banks. DA - 2021/// PY - 2021/// DO - 10.1109/GHTC53159.2021.9612484 SP - 281-288 SN - 2377-6919 KW - Food Insecurity KW - Humanitarian Supply Chain KW - Bayesian Structural Time Series KW - Long Short Term Memory KW - Training Length KW - Expanding and Sliding Window ER - TY - JOUR TI - NEW FINDINGS ON STUDENT MULTITASKING WITH MOBILE DEVICES AND STUDENT SUCCESS AU - Eseryel, U. Yeliz AU - Jiang, Dan AU - Eseryel, Deniz T2 - JOURNAL OF INFORMATION TECHNOLOGY EDUCATION-INNOVATIONS IN PRACTICE AB - Aim/Purpose: This paper investigates the influence of university student multitasking on their learning success, defined as students’ learning satisfaction and performance. Background: Most research on student multitasking finds student multitasking problematic. However, this research is generally from 2010. Yet, today’s students are known to be digital natives and they have a different, more positive, relationship with mobile technologies. Based on the old findings, most instructors ban mobile technology use during instruction, and design their online courses without regard for the mobile technology use that happens regardless of their ban. This study investigates whether today’s instructors and learning management system interface designers should take into account multitasking with mobile technologies. Methodology: A quasi-experimental design was used in this study. Data were collected from 117 students across two sections of an introductory Management Information Systems class taught by the first author. We took multiple approaches and steps to control for confounding factors and to increase the internal validity of the study. We used a control group as a comparison group, we used a pre-test, we controlled for selection bias, and we tested for demographic differences between groups. Contribution: With this paper, we explicated the relationship between multitasking and learning success. We defined learning success as learning performance and learning satisfaction. Contrary to the literature, we found that multitasking involving IT texting does not decrease students’ learning performance. An explanation of this change is the change in the student population, and the digital nativeness between 2010s and 2020 and beyond. Findings: Our study showed that multitasking involving IT texting does not decrease students’ performance in class compared to not multitasking. Secondly, our study showed that, overall, multitasking reduced the students’ learning satisfaction despite the literature suggesting otherwise. We found that attitude towards multitasking moderated the relationship between multitasking and learning satisfaction as follows. Individuals who had a positive attitude towards multitasking had high learning satisfaction with multitasking. However, individuals who had positive attitude toward multitasking did not necessarily have higher learning performance. Recommendations for Practitioners: We would recommend both instructors and the designers of learning management systems to take mobile multitasking into consideration while designing courses and course interfaces, rather than banning multitasking, and assuming that the students do not do it. Furthermore, we recommend including multitasking into relevant courses such as Management Information Systems courses to make students aware of their own multitasking behavior and their results. Recommendation for Researchers: We recommend that future studies investigate multitasking with different instruction methods, especially studies that make students aware of their multitasking behavior and its outcomes will be useful for next generations. Impact on Society: This paper investigates the role of mobile multitasking on learning performance. Since mobile technologies are ubiquitous and their use in multitasking is common, their use in multitasking affects societal performance. Future Research: Studies that replicate our research with larger and more diverse samples are needed. Future research could explore research-based experiential teaching methods, similar to this study. DA - 2021/// PY - 2021/// DO - 10.28945/4723 VL - 20 SP - 21-35 SN - 2165-316X KW - multitasking KW - undergraduate students KW - learning success KW - learning performance KW - learning satisfaction KW - quasi-experiment ER - TY - JOUR TI - Unifying Domain Adaptation and Domain Generalization for Robust Prediction Across Minority Racial Groups AU - Khoshnevisan, Farzaneh AU - Chi, Min T2 - MACHINE LEARNING AND KNOWLEDGE DISCOVERY IN DATABASES AB - In clinical deployment, the performance of a model trained from one or more medical systems often deteriorates on another system and such deterioration is especially evident among minority patients who often have limited data. In this work, we present a multi-source adversarial domain separation (MS-ADS) framework which unifies domain adaptation and domain generalization. MS-ADS is designed to address two types of discrepancies: covariate shift stemming from differences in patient populations, and systematic bias on account of differences in data collection procedures across medical systems. We evaluate MS-ADS for early prediction of septic shock on three tasks. On a task of domain adaptation across three medical systems, we show that by leveraging data from multiple systems while accounting for both types of discrepancies, MS-ADS improves the prediction performance across all three systems; on a task of domain generalization to an unseen medical system, we show that MS-ADS can perform better than or close to the gold standard supervised models built for the system; last but not least, on a task that involves both domain adaptation and domain generalization: generalization to unseen racial groups across medical systems, MS-ADS shows robust out-performance by addressing covariate shift across different racial groups and systematic bias across medical systems simultaneously. DA - 2021/// PY - 2021/// DO - 10.1007/978-3-030-86486-6_32 VL - 12975 SP - 521-537 SN - 1611-3349 KW - Domain adaptation KW - Domain generalization KW - Cross-racial transfer KW - Septic shock ER - TY - JOUR TI - A Theoretical and Evidence-Based Conceptual Design of MetaDash: An Intelligent Teacher Dashboard to Support Teachers' Decision Making and Students' Self-Regulated Learning AU - Wiedbusch, Megan D. AU - Kite, Vance AU - Yang, , Xi AU - Park, Soonhye AU - Chi, Min AU - Taub, Michelle AU - Azevedo, Roger T2 - FRONTIERS IN EDUCATION AB - Teachers’ ability to self-regulate their own learning is closely related to their competency to enhance self-regulated learning (SRL) in their students. Accordingly, there is emerging research for the design of teacher dashboards that empower instructors by providing access to quantifiable evidence of student performance and SRL processes. Typically, they capture evidence of student learning and performance to be visualized through activity traces (e.g., bar charts showing correct and incorrect response rates, etc.) and SRL data (e.g., eye-tracking on content, log files capturing feature selection, etc.) in order to provide teachers with monitoring and instructional tools. Critics of the current research on dashboards used in conjunction with advanced learning technologies (ALTs) such as simulations, intelligent tutoring systems, and serious games, argue that the state of the field is immature and has 1) focused only on exploratory or proof-of-concept projects, 2) investigated data visualizations of performance metrics or simplistic learning behaviors, and 3) neglected most theoretical aspects of SRL including teachers’ general lack of understanding their’s students’ SRL. Additionally, the work is mostly anecdotal, lacks methodological rigor, and does not collect critical process data (e.g. frequency, duration, timing, or fluctuations of cognitive, affective, metacognitive, and motivational (CAMM) SRL processes) during learning with ALTs used in the classroom. No known research in the areas of learning analytics, teacher dashboards, or teachers’ perceptions of students’ SRL and CAMM engagement has systematically and simultaneously examined the deployment, temporal unfolding, regulation, and impact of all these key processes during complex learning. In this manuscript, we 1) review the current state of ALTs designed using SRL theoretical frameworks and the current state of teacher dashboard design and research, 2) report the important design features and elements within intelligent dashboards that provide teachers with real-time data visualizations of their students’ SRL processes and engagement while using ALTs in classrooms, as revealed from the analysis of surveys and focus groups with teachers, and 3) propose a conceptual system design for integrating reinforcement learning into a teacher dashboard to help guide the utilization of multimodal data collected on students’ and teachers’ CAMM SRL processes during complex learning. DA - 2021/2/19/ PY - 2021/2/19/ DO - 10.3389/feduc.2021.570229 VL - 6 SP - SN - 2504-284X KW - self-regulated learning (SRL) KW - teacher decision making KW - learning KW - multimodal data KW - teacher dashboards ER - TY - JOUR TI - Leveraging Granularity: Hierarchical Reinforcement Learning for Pedagogical Policy Induction AU - Zhou, Guojing AU - Azizsoltani, Hamoon AU - Ausin, Markel Sanz AU - Barnes, Tiffany AU - Chi, Min T2 - INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE IN EDUCATION DA - 2021/8/16/ PY - 2021/8/16/ DO - 10.1007/s40593-021-00269-9 SP - SN - 1560-4306 KW - Hierarchical reinforcement learning KW - Decision granularity KW - Pedagogical policy ER - TY - JOUR TI - Forests After Florence: an informal community-engaged STEM research project promotes STEM identity in disaster-impacted students AU - Mulvey, Kelly Lynn AU - Joy, Angelina AU - Caslin, Michael AU - Orcutt, Darby AU - Eseryel, Deniz AU - Katti, Madhusudan T2 - RESEARCH IN SCIENCE & TECHNOLOGICAL EDUCATION AB - Background: Natural disasters, such as hurricanes, can have lasting impacts on a communityPurpose: This research evaluated how participation in an STEM education intervention after an ecological disaster affected students’ persistence, resilience, and STEM identitySample: Hurricane Florence impacted college students (N = 50) were recruitedDesign and Methods: Participants completed pre-test, post-test and daily diary measures before, during and after they completed an intervention where they collected forestry data in their home hurricane-impacted communitiesResults: Participants reported higher STEM identity following the intervention learning experience. Daily interest and enjoyment in science was higher on days when they reported more positive experiences. For resilience, for male students, but not female students, the learning opportunity fostered resilience. Male students reported higher STEM identity on days when they reported more positive learning experiencesConclusion: These findings highlight the benefit of STEM education learning opportunities, particular for disaster-impacted students. DA - 2021/6/22/ PY - 2021/6/22/ DO - 10.1080/02635143.2021.1944077 VL - 6 SP - SN - 1470-1138 KW - Resilience KW - STEM identity KW - gender KW - forestry KW - hurricanes ER - TY - JOUR TI - The Impact of Looking Further Ahead: A Comparison of Two Data-driven Unsolicited Hint Types on Performance in an Intelligent Data-driven Logic Tutor AU - Cody, Christa AU - Maniktala, Mehak AU - Lytle, Nicholas AU - Chi, Min AU - Barnes, Tiffany T2 - INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE IN EDUCATION AB - Research has shown assistance can provide many benefits to novices lacking the mental models needed for problem solving in a new domain. However, varying approaches to assistance, such as subgoals and next-step hints, have been implemented with mixed results. Next-Step hints are common in data-driven tutors due to their straightforward generation from historical student data, as well as research showing positive impacts on student learning. However, there is a lack of research exploring the possibility of extending data-driven methods to provide higher-level assistance. Therefore, we modified our data-driven Next-Step hint generator to provide Waypoints, hints that are a few steps ahead, representing problem-solving subgoals. We hypothesized that Waypoints would benefit students with high prior knowledge, and that Next-Step hints would most benefit students with lower prior knowledge. In this study, we investigated the influence of data-driven hint type, Waypoints versus Next-Step hints, on student learning in a logic proof tutoring system, Deep Thought, in a discrete mathematics course. We found that Next-Step hints were more beneficial for the majority of students in terms of time, efficiency, and accuracy on the posttest. However, higher totals of successfully used Waypoints were correlated with improvements in efficiency and time in the posttest. These results suggest that Waypoint hints could be beneficial, but more scaffolding may be needed to help students follow them. DA - 2021/5/21/ PY - 2021/5/21/ DO - 10.1007/s40593-021-00237-3 SP - SN - 1560-4306 KW - Tutoring system KW - Hints KW - Assistance KW - Data-driven methods ER - TY - JOUR TI - Avoiding Help Avoidance: Using Interface Design Changes to Promote Unsolicited Hint Usage in an Intelligent Tutor (September, 10.1007/s40593-020-00213-3, 2020) AU - Maniktala, Mehak AU - Cody, Christa AU - Barnes, Tiffany AU - Chi, Min T2 - INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE IN EDUCATION AB - A Correction to this paper has been published: https://doi.org/10.1007/s40593-020-00232-0 DA - 2021/3// PY - 2021/3// DO - 10.1007/s40593-020-00232-0 VL - 31 IS - 1 SP - 154-155 SN - 1560-4306 ER - TY - JOUR TI - Predictive Student Modeling in Game-Based Learning Environments with Word Embedding Representations of Reflection AU - Geden, Michael AU - Emerson, Andrew AU - Carpenter, Dan AU - Rowe, Jonathan AU - Azevedo, Roger AU - Lester, James T2 - INTERNATIONAL JOURNAL OF ARTIFICIAL INTELLIGENCE IN EDUCATION DA - 2021/3// PY - 2021/3// DO - 10.1007/s40593-020-00220-4 VL - 31 IS - 1 SP - 1-23 SN - 1560-4306 KW - Student modeling KW - Early prediction KW - Game-based learning environments KW - Self-regulated learning KW - Reflection ER -