Program Sessions

Program Session Descriptions

Welcome and Keynote (9:00 - 10:30)

Welcome - Watch Opening Remarks (opens in YouTube)

Drew Paulin, Academic Director, Data Science Program, School of Information, UC Berkeley

Opening Remarks - Watch Opening Remarks (opens in YouTube)

Catherine Koshland, Vice Chancellor for Undergraduate Education, UC Berkeley

Bio: Catherine P. Koshland is the Vice Chancellor for Undergraduate Education at the University of California, Berkeley, and the Wood-Calvert Professor in Engineering. She is a professor of Environmental Health Sciences in the School of Public Health(link is external)and a professor in the Energy and Resources Group(link is external). Professor Koshland graduated with a B.A. in Fine Arts from Haverford College, studied painting at the New York School of Drawing, Painting and Sculpture, and received her M.S. in 1978 and her Ph.D. in 1985 in Mechanical Engineering from Stanford University. She is a former member of the Haverford College Board of Managers (1994 – 2014; Board Co-Chair 2005 to 2009, Chair 2009 to 2014). More

Keynote- Watch Keynote (opens in YouTube)   View Slides

Tim McKay, Arthur F. Thurnau Professor of Physics, Astronomy, and Education at the University of Michigan


This century opened amidst an information revolution which promises to change higher education as dramatically as industrialization 100 years ago. Many things are already different: information is ubiquitously available, lectures are being flipped, and students are distracted in class by social media. But the real revolution will come when we harness information technology more deliberately to better understand and personalize education. Universities already gather rich and extensive information about each student’s background, interests, behavior, and accomplishments. With care, we can focus attention on measuring what matters for student success and build systems which personalize education at scale. Our goal is to create much greater student motivation and engagement, optimizing our education of an increasingly diverse student body. This work is especially important when we educate at scale, for example in large, foundational courses.

This talk will describe ten years of work in Learning Analytics at the University of Michigan, providing examples of discoveries made, institutional support structures established, and tools developed to help personalize education at scale.

Bio: Prof. Tim McKay is currently an Arthur F. Thurnau Professor of Physics, Astronomy, and Education at the University of Michigan, where his research in astrophysics has focused on the growth and nature of cosmic structure as well as the expansion history of the Universe. His work in Education aims at understanding and improving postsecondary student outcomes using the rich, extensive, and complex digital data produced in the course of educating students in the 21st century. Prof. McKay helped to launch a campus-wide learning analytics effort, chairing the Provost’s Learning Analytics Task Force at UM. In 2014, he became the PI on the REBUILD project, an ongoing college-wide effort to increase the use of evidence-based methods in UM’s introductory STEM courses. He is also currently the director of the Digital Innovation Greenhouse, an education technology accelerator within the UM Office of Digital Education and Innovation.

Morning Sessions (10:45 - 11:45)

Banatao Auditorium  Room 250 Room 254
First Presentation: First Presentation: First Presentation:
Title: Question & Answer with Tim McKay   Title: Naked in the Garden: Shopping for Privacy in the Learning Data Store


Title:  The Average Student Does Not Exist

View Slides


Join Tim for a short question/answer session directly following the keynote.

Presenter(s): Tim McKay, Professor of Physics, Astronomy, and Education, University of Michigan

Topic Area(s): Policy, The LMS and Third Party Tools, Research, Advising, Instruction


If the Learning Management System is academy’s digital garden of knowledge, discourse, and discovery, and the learner’s every insight, assignment, and clickstream in the garden is collected in the Learning Data Store for future analysis and examination, what do we say to students about their privacy? “Grab a fig leaf?”

Instead, let's try on UC Berkeley’s Privacy Balancing Analysis and add a Privacy Interoperability Standard to our wish list.

Presenter(s): Lisa Ho, Campus Privacy Officer, UC Berkeley 

Topic Area(s): Privacy, Instruction


In high-stakes testing, student performance is most commonly judged by one’s relation to the average total score. Data from 1.5k computer science finals graded with Gradescope suggests this may be an ineffective way to characterize student performance.

Presenter(s): Sergey Karayev, Co-Founder, Gradescope

Topic Area(s): Instruction


Second Presentation: Second Presentation: Second Presentation:
Title: Student Responses to Learning Analytics: The Impact of Nods and Nudges

Title: UC’s Learning Data Privacy Principles and Recommended Practices

View Slides

Title: Visualizing Course-Taking Patterns with t-SNE

View Slides


Giving learning analytics directly to students is a highly scalable form of intervention. Further, this approach may promote deeper learning outcomes such as self-regulated learning.  However, some institutions have been hesitant to provide students with these tools out of concern that they would be misinterpreted or potentially demotivating.  In this presentation, we will review the results from two research studies that examine student responses to receiving learning analytics.  The results suggest that students are indeed interested in these tools and that they benefit from them; the results, however, have interesting surprises related to high-achieving students and lead to several open questions inviting further research.

Presenter(s): Mike Sharkey, Vice President for Analytics, Blackboard, Inc.  

Topic Area(s): Policy, Research, Advising, Instruction


Overview of the Learning Data Privacy Principles developed by the University of California Educational Technology Leadership Committee (ETLC). These principles assist campuses and service providers to gather, store, and protect learning data using ethical and secure measures.


Jim Phillips, Director, Learning Technologies, ITS, UC Santa Cruz

Jim Williamson, Director, Campus Educational Technology Systems & Administration, UC Los Angeles

Mary-Ellen Kreher, Director, Course Design and Development, Innovative Learning Technology Initiative, UCOP


Topic Area(s): Privacy, Instruction



Analyzing student course-taking patterns can be complicated. While it's straightforward to group students by ethnicity, major or grade-point average, it's more difficult to identify similar students based on their coursework. Using a machine learning technique called t-sne, I've clustered students based on the courses they took at Berkeley and have been experimenting with ways to visualize the results. The visualization includes color-coding of students by ethnicity, academic division, and time-to-degree.

Presenter(s): Sara Quigley, Data Analyst, Office of Planning & Analysis, UC Berkeley

Topics Area(s): Policy, Instruction, Advising

Birds of a Feather (12:00 - 12:30)

This is an opportunity for participants to talk with each other, and for us to generate ideas and key questions that as a campus community, we want to be thinking about and try to address beyond the conference.

Topics area: Research, Instruction and Advising, Privacy and Policy, Integrating Information Systems/Tools

Locations: TBA

Afternoon Sessions (1:30 - 2:30)

Banatao Auditorium  Room 250 Room 254
First Presentation: First Presentation:
First Presentation:

Title: Earlier and Better Feedback for Student Success: An Analytics Pilot for UC Staff Who Work Closely with Students

View Slides

Title: Moving from a Siloed Data Culture to One of Freedom and Responsibility

View Slides

Title: An Adaptive Equity-Oriented Pedagogy: Using Learning Analytics to Increase Equity and Student Outcomes


The Athletic Study Center (ASC), a campus department that provides academic and advising support to 900 student athletes across 30 teams, is collaborating with Educational Technology Services (ETS) to develop an application that enables ASC staff to identify much earlier in the semester than has previously been possible those student athletes who may be academically at risk. To complement ASC's existing and rich institutional data and tools, the application will provide dynamic indicators on students who may be struggling in the moment. This new application is designed to meet the specific needs of the ASC, with an eye towards eventual adoption by the wider campus community.


Vanessa Kaskiris, Project Manager, Educational Technology Services, UC Berkeley

Flint Hahn, Senior User Experience Designer, Educational Technology Services, UC Berkeley

Topic Area(s): Advising



Seven years after the 1st launch of Cal Answers we find ourselves in an interesting time when we can reflect on lessons learned and plan our next steps in advancing the campus data culture.  In this session we will cover our lessons learned in data warehousing as well as the challenges in combining traditional EDW data with new and dynamic learning analytics data. We will end with an audience discussion about how to advance the campus data culture.   We look forward to an engaged discussion with you.


Amber Machamer, Executive Director, Office of Planning and Analysis, UC Berkeley

Mark Chiang, Enterprise Data Warehouse Manager, Information Services & Technology - Enterprise Data, UC Berkeley

Topic Area(s): Policy, Research



We will present an adaptive equity-oriented pedagogy (AEP), which is a teaching approach where instructors use learning analytics from students’ diagnostic assessments and engagement data to improve academic achievement. We will explain how instructors enhanced student learning via AEP by leveraging insights from LMS analytics on weekly assessments and engagement with instructional materials (e.g., readings, flipped classroom videos, etc.). AEP  students outperformed a control group by a letter grade on average (p<.0001), instructors received significantly higher course ratings, and students demonstrated greater improvements in psychosocial outcomes-- i.e., reduced stereotype threat and increased motivation and confidence. Our learning analytics data shows how these findings held when accounting for students’ race, gender, disability, and sexual orientation.


Andrew Estrada Phuong, Chancellor’s Fellow, Consultant, and Program Developer, Division of Equity and Inclusion, UC Berkeley

Judy Nguyen, Graduate Student and Researcher, Graduate School of Education, Stanford University

Ashley Lopez, Research Manager, Design for Equity Lab in D-Lab, UC Berkeley

Danielle Hoague, Research Manager, Design for Equity Lab in D-Lab, UC Berkeley

Diana Heath, Research Manager, Design for Equity Lab in D-Lab, UC Berkeley

Bernadette Blashill, Researcher, Design for Equity Lab in D-Lab, UC Berkeley

Shahana S. Farooqi, Instructor and Researcher, UC Berkeley

Claire S. Bang, Instructor and Researcher, UC Berkeley

Heaven Taylor, Researcher, Design for Equity Lab in D-Lab, UC Berkeley

Estefania Laines, Researcher, Design for Equity Lab in D-Lab, UC Berkeley

Negeen Khandel, Researcher, Design for Equity Lab in D-Lab, UC Berkeley

Topic Area(s): Research, Instruction

Second Presentation:
Second Presentation:
Second Presentation:

Title: UC Berkeley's Learning Data in Action

View Slides

Title: Reasoning About Knowledge From Learner Pathway Information

Capturing Student "Impact" in Collaborative Online Learning

View Slides


ETS is developing an application to help tutors and advisors identify at-risk students within cohorts (as opposed to within an individual course). The current proof of concept will be demoed, and the challenges of marshaling the necessary underlying data will be discussed, along with some of the powerful implications this work has for a democratized approach to campus data access and use.


Sandeep Jayaprakash, Data Scientist/Applications Programmer, Educational Technology Services, UC Berkeley

Ray Davis, Applications Programmer, Educational Technology Services, UC Berkeley

Topic Area(s): Data Processing Infrastructure, Advising


The aggregate behaviors of students in a digital learning environment can collectively encode information about the pedagogical objects with which they interact. In this presentation, I will demonstrate ways in which the synthesis of these data, from K-12 systems and higher-ed, can illuminate the terrain of their educational environment and support them in their decision making and wayfinding. A novel application of representation learning, an approach recently popularized by its application to modeling language, is brought to bear on the contexts of student problem solving and course enrollment sequences to create vector representations of these objects. With course representations in hand, the similarity and compositionality of academic units at UC Berkeley is explored. A developing application of this research to practice will be previewed in a system which pulls from multiple sources of information (e.g. registrar’s suggested courses, web-parsed degree-requirements, and currently offered classes) enabling students to form queries that can satisfy complex constraints and factor in features of courses suggested by the representations. Lastly, the utility of guiding students’ traversal between and within courses based on their personal history of interactions is considered.


Zachary Pardos, Assistant Professor, Graduate School of Education, and iSchool, UC Berkeley

Topic Areas: Research, Policy, Advising  


Increasing opportunities for meaningful student collaboration and social interaction have been shown to be significant factors in student success in online learning courses. However, it can be difficult for instructors and students to efficiently facilitate and evaluate peer-to-peer and team-based learning activities. This presentation focuses on a new learning tool called the Impact Studio, a student analytics dashboard that tracks and visualizes student social participation in bCourses Discussions and the  SuiteC collaborative learning software. Uniquely, the Impact Studio reports seven Impact metrics, which represent how a student's contributions are taken up and responded to by the community. We explore student and instructor pilot usage of the Impact Studio (released July, 2017) in an online undergraduate course, the pedagogy underpinning its inclusion in the learning management system, and the utility of Impact metrics as predictors of student success.


Glynda Hull, Professor Graduate School of Education, UC Berkeley

John Scott, GSI/GSR, Graduate School of Education, UC Berkeley

Topic Area(s): LMS and Third Party Tools, Research, Instruction

Berkeley Panel (2:45 - 4:00)

Watch Berkeley Panel (opens in YouTube)   View Slides


Zachary Pardos, Assistant Professor, Graduate School of  Education, and iSchool, UC Berkeley


Jenn Stringer, Assistant Vice Chancellor for Teaching & Learning, and Chief Academic Technology Officer, UC Berkeley

Derek Van Rheenan, Associate Adjunct Professor and Director, Cultural Studies of Sport in Education, UC Berkeley

Anne Baranger, Adjunct Professor, Director of Undergraduate Chemistry, and Faculty Assistant for Teaching and Learning, UC Berkeley

Glynda Hull, Professor, Graduate School of Education, UC Berkeley