The Next Frontier of AI-Enabled Collaborative Learning
Carolyn Rosé, Carnegie Mellon University
November 18, 2020
Dr. Carolyn Rosé joined NYU-LEARN for a cutting-edge talk on how language technologies can support student teamwork and on-the-job workplace learning. Using Online Mob Programming (OMP), Dr. Rosé and her team aim to encourage project teams in the classroom to reflect on concepts and share work in the midst of their project experience with the goal of bringing dynamic collaboration support into physical spaces through Virtual Humans.
Presentation Slides coming soon.
What Data Would Make You Change (Your) Course?
An interactive workshop for NYU Faculty cohosted with NYU Teaching and Learning with Technology. This workshop focused on practical approaches to improving teaching and course design using data. Instructors shared their experiences teaching their classes this (unusual) year, and learned some easy methods of using data to shape a potential improvement into a real plan.
Multimodal Data Stories to Support Reflection in Clinical Simulations
Roberto Martinez-Maldonado, Monash University
October 14, 2020
Dr. Roberto Martinez-Maldonado joined NYU-LEARN for an innovative talk on the use of sensing technologies in clinical simulations. Using co-designed Data Stories and Multimodal Learning Analytics (MMLA) Dr. Martinez-Maldonado and his team aim to provide automated feedback to healthcare students and facilitators that can inform the clinical debrief and help identify areas that need further development.
College in the Time of Corona: Spring 2020 Student Survey Results & Community Discussion
Alyssa Wise and Yoav Bergner, New York University
September 30, 2020
The Learning Analytics Research Network (LEARN) hosted a timely conversation that explored how students across the US experienced the rapid transition to remote instruction in Spring 2020. Drs. Alyssa Wise and Yoav Bergner shared the results from LEARN’s recent IRB-approved, 15 question student survey. Over 250 students from various universities responded to the survey between late March and early May 2020, offering unique insight into the experience of this transition as it occurred. They also lead a discussion of their implications for our teaching and learning practices moving forward.
Tracing Professional Identity Development through (Mixed-Methods) Data Mining
Sameen Reza, New York University
June 18, 2020
As part of ICLS (the International Conference of the Learning Sciences) 2020, we invited members of both the Learning Sciences and Learning Analytics communities for a unique conversation exploring the promise and potential peril of data-mining methods for generating insight into complex human learning processes. The session was built around the accepted ICLS paper “Tracing Professional Identity Development through Mixed-Methods Data Mining of Student Reflections”, with an invited commentary by Professor David Shaffer followed by an open community discussion.
Presentation Slides on Tracing Professional Identity Development through (Mixed-Methods) Data Mining Video Recording of Tracing Professional Identity Development through (Mixed-Methods) Data Mining presentation
Learning Analytics: Facing Up to Ethical Challenges
Rebecca Ferguson, The Open University
March 3, 2020
An insightful talk into the ways in which The Open University, UK is working to build a culture of responsible learning analytics use. Dr. Rebecca Ferguson described six key ethical challenges and how they are currently being tackled at The Open University, UK.
Debating Data Dilemmas
A Workshop by NYU-LEARN
February 18, 2020
How do we balance insight into learning with protecting student privacy? What are the implications of using analytics to designate certain students as “at-risk”? Should I be able to compare my data to other students’? In this workshop, the NYU-LEARN team facilitated discussion and debate of real world dilemmas that have arisen for the use of analytic data to support learning. Talking in small groups, participants identified the different principles and tensions at play from different perspectives. Through the event, participants developed a deeper understanding of the complex balances at play in learning analytics work and how to support capacity-building efforts to address them at New York University.
Data Enhanced Learning and the Quantified Student
Marcus Specht, Delft University of Technology
November 21, 2019
Social and educational sciences have been collecting data about the effectiveness and the efficiency of educational practices for some time. Especially in the last years more data driven approaches have been developed in the field of Learning Analytics to create dashboards and tools for monitoring educational activities. Reflection and personalised feedback are some of the most efficient means for enhancing human learning. Without feedback and making our progress visible and tangible we are lost in our way. Data tracking in digital systems and sensor-based wearable computing enable data-informed learning support for reflection and personalisation. In this talk, Dr. Specht shared the work he and his team are doing to develop data-enhanced learning and human-centred learning analytics at the Leiden-Delft-Erasmus Center for Education and Learning.
Developing 21st-Century Skills with Multimodal Analytics Interactive Demos and Discussion
Xavier Ochoa, New York University
November 12, 2019
An interactive experience and exploration of how Multimodal Learning Analytics (MmLA) can improve learning and teaching processes. Xavier Ochoa, one of LEARN’s core faculty and Assistant Professor of Learning Analytics, offered hands-on demonstrations of two cutting-edge systems for collecting real-time data in physical spaces to automatically generate feedback for communication and collaboration skills. Demos were followed by a discussion of the ways these tools can be applied to support students and instructors.
SoLAR Webinar: Designing Learning Analytics for Humans with Humans
Alyssa Wise, New York University
October 16, 2019
To be effective learning analytics tools must not only be technically robust but also designed to support use by real people. In this webinar, Alyssa Wise presented a diverse set of examples of the ways that NYU-LEARN is including educators and students in the process of building and implementing learning analytics. She shared examples of how to: involve students in the creation and revision of learning analytics solutions for their own use; work with instructors to align analytically available metrics with valued course pedagogy; and partner with an educational team to design and implement interventions based on at-risk students predictions. Watch this webinar to gain a sense of both the conceptual issues and practical concerns involved in designing learning analytics for humans with humans.
Panel: Data Analytics and Student Success
Xavier Ochoa, New York University
October 15, 2019
“Big Data” analytics have permeated many aspects of our lives, for good and ill. This discussion centered around how educational institutions leverage the power of data analytics to improve student learning, persistence, and graduation. Panelists also discussed how to gain these benefits without violating privacy, marginalizing teachers, or introducing systematic bias that may harm some students.
Learning Analytics for Lectures, Design and Informal Learning: Innovations from Japan
October 3, 2019
A series of flash talks from our visitors Drs. Hiroaki Ogata, Atsushi Shimada and Masanori Yamada from Kyoto University in which they describe their work exploring the possibilities for learning analytics in face-to-face lectures, informal learning environments, and to inform course design.
Connecting Formal and Informal Learning through Learning Evidence and Analytics, Hiroaki Ogata
Toward Learning Analytics for Reconsideration of Instructional Design, Masanori Yamada
Advanced Learning Analytics for Face-to-face Lecture Support, Atsushi Shimada
Aligning Learning Analytics with Classroom Practices & Needs
Simon Knight, University of Technology Sydney
October 1, 2019
How do we make use of data about our students to support their learning while ensuring that learning analytics developers and educators to align with educator practice and needs? The University of Technology Sydney has taken a participatory design based approach to designing and implementing learning analytics in practice, and understanding their impact. Their work has identified existing practices with which learning analytics may be aligned to augment them. In this talk, Dr. Knight introduced some of these projects, particularly drawing on their work in developing analytics to support student writing (writing analytics), giving examples of how analytics were aligned with existing pedagogic practices to support learning. Through this augmentation, supported by design-based approaches, Dr. Knight argues they can develop research and practice in tandem.
Advanced Analytics in Medical Education
Martin Pusic, NYU Langone Health
April 16, 2019
New educational technologies enable innovative health professions approaches with the potential to promote patient safety, better focus on learner needs, and enhance learning efficiency and effectiveness. In essence, the goal is to help learners climb the learning curve to expert clinical performance faster without sacrificing long-term retention. Dr. Pusic described a novel training approach that uses hundreds of Cognitive Simulations of radiology and ECG cases. This system not only provides an immense volume of simulated cases, but also allows strategic selection of cases to optimize learning according to the learner’s current skill state. Using this approach, learners can achieve in hours or days a level of experience and performance that would normally require years to accumulate. The vision is to have learners practice on simulated cases in the same way a violinist would practice their scales: practicing to mastery before attempting the performance that matters (i.e., in a real-life setting). Emerging evidence suggests important differences in the number, type, and sequence of cases, and the manner in which feedback is provided. Spoiler alert: Training to expertise probably requires far more practice than is currently done.
From Attendance to Analytics: Traditional Strategies & Emerging Approaches to Data-Informed Teaching
A panel hosted in collaboration with FAS Office of Educational Technology.
March 26, 2019
Instructors routinely adjust their courses based on diverse forms of data, from attendance records to quiz grades to student annotations on readings. Faculty panelists explored a variety of ways faculty do and could use data to inform their teaching by sharing their experiences and methods. Experts in learning research discussed ways to enhance these practices with services such as the NYU Learning Analytics Dashboard.
Directions for Making the Most of Learning Analytics
Dragan Gašević, Monash University
March 12, 2019
The analysis of data collected from user interactions with technology has attracted much attention as a promising approach to enhancing the human learning process. This growing interest led to the formation of the field of learning analytics. The field has now entered the next phase of maturation with a growing community who has an evident impact on research, practice, policy, and decision-making. This talk first provided a brief overview of recent developments in the field and then explore two key challenges for learning analytics that require immediate attention: i) validity of data collection and analysis and ii) user interaction with results of analytics to inform action. The talk shared promising directions for addressing the two challenges by considering learning analytics as an interdisciplinary interplay between data science, theories of human learning, and design.
Automated Writing Evaluation Feedback to Support Learning
Jill Burstein, Educational Testing Service
February 12, 2019
Automated writing evaluation (AWE) uses natural language processing (NLP) methods to detect and extract linguistic features relevant to writing quality (for example use of sources, claims, and evidence; topic development; coherence; and grammatical conventions). The talk described the inner-workings of AWE systems – specifically, what these systems can do now. This was illustrated with a demo of Writing MentorTM (WM) - a Google Docs add-on that provides students with AWE-based feedback to help them improve their writing in a principled manner - and the presentation of results from qualitative usability evaluations with users in the wild.
Learning from Every Student: Leveraging Analytics to Support Student Success In Higher Education
Stephanie Teasley, University of Michigan
November 27, 2018
The research community for Learning Analytics is growing rapidly promising new insights into learning and resulting innovations in pedagogy. For this promise to be realized, however, we need the capacity to leverage educational data for scholarly research, and apply research results at the kind of scale that truly changes how we teach and learn. In this talk Dr. Teasley presented how the University of Michigan is engaged in learning analytics as an institutional initiative aimed at leveraging the data produced by digitally-mediated educational tools to better understand and improve student outcomes. Dr. Teasley shared two examples of the way the University of Michigan is using learning analytics to support student success. In the first she described how analytics help us understand the ways curricular pathways impact students. In the second she described our efforts to understand how student-facing dashboards can be designed to support important meta-cognitive skills.
LEARN Lounge at NYU Technology Summit
November 14, 2018
Gesture Detection System & Dynamic Network Wall Demo
Yoav Bergner and Xavier Ochoa
Data-Driven Decision-Making & Learning Design Workshop
Alyssa Wise, Ben Maddox and Yoav Bergner
How exactly can learning analytics inform teaching and course design? Using concrete examples, participants explored the different kinds of insight innovative educational data analysis techniques offer and considered their specific application to our their courses.
Opposites Attract: Analytics and the Humanities Seminar
Robert Squillace and Andrew Brackett
The session exploref analytics use from the instructor’s perspective, focusing on a partnership to develop a faculty-facing analytics tool for a seminar that emphasizes project-based learning. Visual examples illustrated how the learning analytics service supported instructional decision-making and showcase recent enhancements to the data visualizations.
More Than Lightning in a Bottle: (How) Will Learning Analytics Transform Teaching & Learning?
Alyssa Wise, Martin Pusic and Xavier Ochoa
Increased data availability and new analysis methods offer exciting opportunities to look inside learning processes in ways never before possible. But (how) will this transform university teaching and learning?
Quantitative Ethnography: Turning Big Data into Real Understanding
David Williamson Shaffer, UW-Madison School of Education
October 30, 2018
In the age of Big Data, we have more information than ever about what students are doing and how they are thinking. However, the sheer volume of data available can overwhelm traditional qualitative and quantitative research methods, leading to research that finds significance without meaning. The science of quantitative ethnography connects the study of culture with statistical tools to understand learning, taking a critical step in the new field of learning analytics: going beyond looking for patterns in mountains of data to tell textured stories at scale.
Interactive Workshop on Quantitative Ethnography: Open Source Tools for Analyzing Large Sets of Discourse Data
David Williamson Shaffer, UW-Madison School of Education
October 29, 2018
This workshop introduced participants to Quantitative Ethnography, a set of tools for modeling complex and collaborative thinking. A central premise of Quantitative Ethnography is learning is a process of enculturation in which students learn to make relevant connections among the skills, concepts, and/or practices in a domain. Quantitative Ethnography models the structure of these connections in large- and small-scale datasets, and logfiles of many kinds, including transcripts of structured and semi-structured interviews or video data, games and simulations, chat, email, and social media. By modeling patterns of connections in discourse, Quantitative Ethnography helps researchers quantify and visualize the development of complex and collaborative thinking.
This interactive workshop provided an overview of Quantitative Ethnography, with an emphasis on the conceptual and practical issues of data management, coding, and modeling, and open-source tools to address these issues: nCoder, a tool for generating and validating qualitative codes; and ENA, a tool for modelling, visualizing, and testing connections in data.
Learning Analytics in Physical Spaces: Capturing & Analyzing Multimodal Learning Traces
Xavier Ochoa, New York University
October 2, 2018
The goal of Learning Analytics is to understand and improve learning. However, learning does not always occur mediated by a computational system. It also happens in face-to-face, hands-on, unbounded and analog learning settings such as classrooms and labs. The sub-field of Multimodal Learning Analytics (MMLA) emphasizes the analysis of natural rich modalities of communication and expression during learning activities such as students’ actions, speech, writing, and nonverbal interaction (e.g. gestures). This talk explored several techniques to capture and analyze multimodal learning traces, giving examples of how to use these multimodal analyses to understand the learning process in physical contexts and provide feedback to its participants. The talk concluded with a discussion of opportunities that Learning Analytics opens for non-digital learning.
Reflective Writing Analytics
Andrew Gibson, Research Fellow in Writing Analytics at the Connected Intelligence Centre, University of Technology Sydney
March 8, 2017
Reflective Writing Analytics (RWA) brings together two worlds: the human world of reflection, and the computational world of analysis. RWA holds potential to discover latent themes, temporal patterns, and features related to the well-being of the writer. However, RWA is complicated by explanatory divergence between its psychosocial and computational dimensions, making it difficult to compute analytics that are considered meaningful to the people that use it.