More Like this

1. Multimodal analytics for collaborative teacher reflection of human-ai hybrid teaching: design opportunities and constraints

   https://doi.org/10.1007/978-3-031-42682-7_45  

   Karumbaiah, S.; Liu, P.; Maksimova, A.; De Vylder, L.; Rummel, N.; Aleven, V. (August 2023, Springer)

   Viberg, O.; Jivet, I.; Muñoz-Merino, P.; Perifanou, M.; Papathoma, T. (Ed.)

   Past research shows that teachers benefit immensely from reflecting on their classroom practices. At the same time, adaptive and artificially intelligent (AI) tutors are shown to be highly effective for students, especially when teachers are involved in supporting students’ learning. Yet, there is little research on how to support teachers to reflect on their practices around AI tutors. We posit that analytics built on multimodal data from the classroom (e.g., teacher position, student-AI interaction) would be beneficial in providing effective scaffolding and evidence for teachers’ collaborative reflection on human-AI hybrid teaching. To better understand the design opportunities and constraints of a future tool for teacher reflection, we conducted storyboarding sessions with seven in-service teachers. Our analysis revealed that certain modalities (e.g., position v. video) might be more beneficial and less constrained than others in identifying reflection-worthy moments and trends. We discuss teachers’ needs for reflection in classrooms with AI tutors and their boundaries in using multimodal analytics. 

   more » « less
2. Design in the HCI Classroom: Setting a Research Agenda

   https://doi.org/10.1145/3322276.3322381  

   Wilcox, Lauren; DiSalvo, Betsy; Henneman, Dick; Wang, Qiaosi (June 2019, Proceedings of the 2019 ACM Conference on Designing Interactive Systems)

   Interaction design is playing an increasingly prominent role in computing research, while professional user experience roles expand. These forces drive the demand for more de- sign instruction in HCI classrooms. In this paper, we distill the popular approaches to teaching design to undergraduate and graduate students of HCI. Through a review of existing research on design pedagogy, an international survey of 61 HCI educators, and an analysis of popular textbooks, we explore the prominent disciplinary perspectives that shape design education in the HCI classroom. We draw on our analyses to discuss the differences we see in forms of design taught, approaches to adapting design instruction in computing-based courses, and the tensions faced by instructors of these classes. We conclude by arguing for the importance of pedagogical research on design instruction as a vital and foundational area of inquiry in Interaction Design and HCI. 

   more » « less
3. Using Vignettes to Categorize Behaviors That Students Associate with Dispositions

   https://doi.org/10.1109/FIE61694.2024.10892902  

   Sabin, Mihaela; McCauley, Renée; MacKellar, Bonnie; Kumar, Amruth N (October 2024, IEEE)

   This full research paper contributes to current work on fostering professional dispositions in computing and engineering education by identifying the categories of behaviors that students associate with dispositions while doing course work. Professional dispositions, demonstrated through desirable behaviors in the workplace, such as being persistent or self-directed, are explicitly sought by employers. Fostering dispositions among students has been identified in various curricular recommendations as an important goal. In prior work, the authors used reflection exercises, in which students were presented with the definition of a disposition and asked to answer an open-ended reflection prompt on how they applied the disposition in their own work. Thematic analysis of student responses to reflection exercises resulted in categories of behaviors that students associated with dispositions. In the work discussed in this paper, the authors used vignette exercises to collect and analyze similar data and gain further insight into behavioral categories and students’ perceptions of dispositions. Vignettes include short scenarios that demonstrate the application of dispositions in real life. A vignette exercise involves students reading a vignette scenario, identifying the disposition demonstrated by the scenario, and answering the same open-ended reflection prompt as in the reflection exercises from the earlier studies. The research question for this study is: Which behavioral categories obtained from analyzing student responses to reflection exercises were confirmed using vignette exercises (and which were not confirmed), and which behavioral categories were refined? To answer this question, researchers from four different institutions of higher education collected data in multiple courses over two semesters. The student open-ended responses to vignettes were thematically analyzed to identify behavioral categories for four dispositions: collaborative, meticulous, persistent and self-directed. The ultimate goal of this work is to create classroom interventions and learning activities that foster dispositions among students based on behavioral categories. This study supports this goal in two ways. It provides another iteration of behavioral category analysis and introduces vignettes to encourage students to reflect candidly and communicate clearly how they apply dispositions in terms of behaviors. The study results and their implications for fostering dispositions in a classroom setting are presented and discussed. 

   more » « less
4. Leveraging Affect Transfer Learning for Behavior Prediction in an Intelligent Tutoring System

   https://doi.org/10.1109/FG52635.2021.9667001  

   Ruiz, Nataniel; Yu, Hao; Allessio, Danielle A.; Jalal, Mona; Joshi, Ajjen; Murray, Thomas; Magee, John J.; Whitehill, Jacob R.; Ablavsky, Vitaly; Arroyo, Ivon; et al (December 2021, 16th IEEE International Conference on Automatic Face and Gesture Recognition (FG 2021))

   In this work, we propose a video-based transfer learning approach for predicting problem outcomes of students working with an intelligent tutoring system (ITS). By analyzing a student's face and gestures, our method predicts the outcome of a student answering a problem in an ITS from a video feed. Our work is motivated by the reasoning that the ability to predict such outcomes enables tutoring systems to adjust interventions, such as hints and encouragement, and to ultimately yield improved student learning. We collected a large labeled dataset of student interactions with an intelligent online math tutor consisting of 68 sessions, where 54 individual students solved 2,749 problems. We will release this dataset publicly upon publication of this paper. It will be available at https://www.cs.bu.edu/faculty/betke/research/learning/. Working with this dataset, our transfer-learning challenge was to design a representation in the source domain of pictures obtained “in the wild” for the task of facial expression analysis, and transferring this learned representation to the task of human behavior prediction in the domain of webcam videos of students in a classroom environment. We developed a novel facial affect representation and a user-personalized training scheme that unlocks the potential of this representation. We designed several variants of a recurrent neural network that models the temporal structure of video sequences of students solving math problems. Our final model, named ATL-BP for Affect Transfer Learning for Behavior Prediction, achieves a relative increase in mean F -score of 50 % over the state-of-the-art method on this new dataset. 

   more » « less
5. Problems of Problem-Based Learning: Exploring Meta-Agency in Problem-Based Cybersecurity Learning in College Education

   Park, Jongchan; Deng Yuli; Agrawal Garima; Techawitthayachinda Ratrapee; Chen Ying-Chih; Huang Dijiang; Liu Huan (April 2023, Annual Meeting of the American Educational Research Association 2023)

   In college cybersecurity education, problem-based learning has been introduced to promote student agency in solving a complex problem. However, a dilemma of balancing the student agency persist and previous research has focused on students’ cognitive, metacognitive, and regulatory to enhance the efficacy of PBL. Given the importance of students’ self-awareness of their agency, this study suggests a concept of meta-agency as an essential learner characteristic that influences the effectiveness of student agency in PBL. Four dimensions of meta-agency, perceptions of productive struggle, expectation alignment between instructor and students, strategies for regulating agency, and familiarity with PBL tasks, were qualitatively explored with student interview data. Features of meta-agency and how students’ meta-agency level develop through cybersecurity PBL sessions were further investigated. 

   more » « less