Educause day 1
Opening ceremony, lots of lights and sparkle like the Emmys. Keynote speaker Sir Ken Robinson. Leading a culture of innovation. We see the world through cultural filters. A a result, trying to predict the effect of IT on culture is impossible until it changes culture. Examples:TV, iPhone, phonograph. Unlike animals we are constantly evolving culturally. Driven by technology. Thus we need to think differently about education. Education still locked into old cultural norms. Old model of college- job set for life is over. Thus so is k-12 standardized testing get into college model. IT the solution?
“OAAI: Deploying an Open Ecosystem for Learner Analytics” Josh Baron, Marist College. The Open Academic Analytics Initiative (OAAI), an NGLC grant recipient, has developed a predictive model for learner analytics using open-source tools, which they are releasing under an open-source license. They share project outcomes along with research into effective OER-based intervention strategies and other critical learner analytics scaling factors. Student attitude data (sat, GRE) other demographics, Sakai lms data as input. Tools: Pentaho Business intelligence oaai predictive model. Predictive modeling markup language. Instructor received academic alert report 3 times during semester. Each with a confidence score. Lms data normalized by class averages. Most powerful predictor is GPA followed by grade book. Number of logins not so much. Contrast to Chico below. Intervention strategies: Awareness note to students; suggest online academic support environment or Learner-facilitator interaction. Both interventions scored same, but better than control. So just simple awareness of risk is enough. But students who received intervention more likely to withdraw, too. Slides at OAAI+EDUCAUSE+Presentation+DRAFT+v1.pptx 3 MB, Powerpoint Slides Uploaded on 10/17/2013 also see solarresearch.org.
“Leveraging Learning Analytics to Build a Culture of Evidence-Based Decision Making” Mesa community college. Speakers : Philip Adams, Director of Client Services
Blackboard Inc. ; Craig Jacobsen Residential Faculty Mesa Community College ;James Mabry Vice President of Academic Affairs Mesa Community College Shouan PanPresident
Community College ; Sasan Poureetezadi Vice President, Information Technology Mesa Community College
See ECAR readiness index http://www.educause.edu/library/resources/2012-ecar-study-analytics-higher-education
Buy in from bottom up and Top down, great support from president. Finally brought In blackboard analytics (The full Learn+SIS version) to pull in all the data. Many levels of access to protect privacy, role based. Bb analytics different from analytics for Learn. Faculty can look at data too.
Advances in Devices, Cloud Services, and Data Analytics: High-Impact Opportunities for Education Anoop Gupta
Microsoft Corporation ,
We are witnessing rapid advances and innovation in devices, complemented by tremendous innovation in cloud services and analytics, which open up new models for scale learning such as MOOCs. New models and pedagogies for in-classroom learning are transforming the nature of textbooks. Novel communication and collaboration services, always-on feedback loops, and big data can offer deep insights into student learning, provide personalized pathways for students, and improve retention. This talk will share a perspective on these ongoing developments and touch on key products, services, and research efforts that can empower the higher education community. Lectures are a terrible way to teach. Students learn <30% of what they don't already know. Almost any type of tech works better, and devices generate numbers that can transform via analytics. Technology adoption phases 1) reproduce analog 2) value added versions. 3) missed it. Moocs gaining acceptance. MOOCs evolving for industrial learning. Next step MEC Massively Empowered Classrooms. Bodies of course materials customized for local delivery. Future: empowered local faculty; modular content with clear rights. Powerful authoring and compositing tools and mashups. Non siload analytics, social tools to keep student engaged, useful credentials
Do Clicks Count to Increase Student Achievement? Learner Analytics on a Large-Enrollment Hybrid Course Kathy Fernandes, Director, Learning Design and Technologies ,California State University: John Whitmer, Program Manager, Academic Technology and Analytics ,California State University
Univ California Chico case study research project. Mining data from SIS and LMS (bb). Used excel, tableau, stata, spss. Looked at one large(373 students) course, biggest raw usage of lms entire campus. 10% increase in mastery! But 7 to 11% increase in DWF. Tinyurl.com/clicksedu13. Early questions what data do we have have and how good is it. lMS usage was 4 times better than sis data in predicting final grade. Lms explained d18% student characteristics 4%. Student demographic data may be too course grained.
Educause day 2
Speeding Up on Curves Bradley Wheeler Vice President for IT and CIO Indiana University WINNER: 2013 EDUCAUSE Leadership Award. Speeding Up On Curves 8 MB, Powerpoint Slides Uploaded on 10/18/2013
At least 3 Curves in the it road:
1.Public to Private Good …(price/cost pressures). Higher Education in Increasingly a Private Good. Colleges and Universities Must Substantially Change our Cost Structure
2.Digital Favors Scale …(rise of substitutes?) 1. $$$ Residential Education (flipped classroom) 2. $$ Online Courses / Degrees 3. $ Massive Online Courses (MOCs) 4. FREE Massive Open Online Courses (MOOCs)
Campus strategies Independence, dependence and interdependence. Scale of digital leads to more options for price conscious students. Upper level courses subsidized by large enrollment freshmen courses.
3.Campus CIO Influence …(who’s driving?) He is very worried.
Improving Student Success: Using Groundbreaking Analytics and Fast Data to Improve Student Retention Vince Kellen, Senior Vice Provost for Academic Planning, Analytics & Technologies University of Kentucky Using Groundbreaking Analytics and Fast Data … 7 MB, Powerpoint Slides Uploaded on 10/24/2013
Take a lecture video, convert speech to text, index keywords to time line, add assessment tool, measure mastery. First steps over the past year
•Mobile micro-surveys: Learning from the learner
•Student enrollment, retention, demographics, performance, K-Score, facilities utilization, instructor workload and more
•High speed, in-memory analytics architectural differences
•Open data and organizational considerations
Coming down the road?
•Micro-segmentation tool to enhance user and IT productivity, develop personalized mobile student interaction/intervention
•Models for learner technographics, psychographics, in addition to behaviors, performance, background
•Advanced way-finding for streaming content like lecture capture
•Content metadata extraction and learner knowledge discovery
•Real-time measures of concept engagement and mastery
•Real-time learner recommendations and support engine
•Use graphing algorithms to perform more sophisticated degree audit what ifs
UK mobile micro survey helps collect data. Enrollment retention and graduation student performance academic career productivity measured as classroom utilization . Thief micro survey app is 97 percent iOS. Collect stats on devices, push alerts, club participation. Tied into blackboard to receive academic warning alerts. Thinking about calendar alarm alerts. Goo to class wake up. MOOKs let’s take aiming course and make it bigger. Current products aimed at big enrollment, not so much for upper level oddball courses. Adaptive learning and metadata system Learns keywords and map to course topics. Common theme use big data to identify either at risk or even concept level difficulties and apply some automatic intervention. Major institutional commitment to aggregator data silos into open database and hire 3 ph.d statisticians. Raise skill set in colleges and units. Totally revs up IR department. Open data to whatever tools end units want sas spss r.
This years theme is connected connections. To that end, presenting “Higher Education Is a Massively Multiplayer Game,” Jane McGonigal, game designer and author who advocates the use of mobile and digital technology to channel positive attitudes and collaboration in a real world context. She has taught game design and game studies at the San Francisco Art Institute and the University of California, Berkeley, and currently serves as the Director of Game Research & Development at Institute for the Future  and Chief Creative Officer at SuperBetter Labs.
In the best-designed games, our human experience is perfectly optimized: we have important work to do, we’re surrounded by potential collaborators, and we learn quickly and in a low-risk environment. When we’re playing a good online game, we get constant useful feedback, we turbocharge the neurochemistry that makes challenge fun, and we feel an insatiable curiosity about the world around us. None of this is by accident. Game developers have spent three decades figuring out how to make us happier and more collaborative, how to make learning more fun and social, and how to satisfy our hunger for meaning and success. All of these game-world insights can be applied directly to reinvent higher education as we know it. Games taken seriously as future of education. Over 1 billion people play at least one hour a day. Average player on “Call of Duty” spends a month of full time work per year playing the game. 81% of US workers show up unmotivated to work. Similar for schools. So they find engagement in games. 10 positive emotions felt by gamers: joy relief love surprise pride curiosity excitement awe/wonder contentment creativity. massively multiplayer thumb wrestling whole hall ~5000 people stood up and played. 3 positive emotions for every one negative breeds a sense of contentment resilience and motivation we wish occurred during learning. The opposite of play is not work it is depression. Playing games light up caudate hippocampus thalamus. Examples foldit http://fold.it/portal/ protein folding gamers better than computers. Evoke a crash course in solving world problems http://www.urgentevoke.com . Find the future the game. http://exhibitions.nypl.org/100/digital_fun/play_the_game
Sherpa: Combine Predictive Analytics with a Recommendation Engine to Improve Student Success Robert Bramucci Vice Chancellor, Technology & Learning Services
South Orange County Community College District http://www.slideshare.net/jpgaston/sherpa-overview-12854534
Recommendations systems, like. Amazon or Netflix personalized unlike most academic systems, Sherpa looks at student profile ando recommends coursesservices information tasks. Triggers generated nudges via portal email txt. Allows student feedback to measure effectiveness. Nudge is personal timely relevant and actionable, not a campus wide broadcast. Predictive analytics. Early warnings, recommendations. Only 15% of their students use LMS so not a rich source. Most systems rely on regression. Computation heavy, data must match assumptions they took a machine intelligence. Worked with guy from Marist. Called in experts. Three math models. And a sea of missing data. Divide data into training and predictive of past to test before predicting future
Adaptive learning back on the rise. Subset of personalized learning.
Educause day 3
From IT to Academic Affairs: Getting Started with Learning Analytics
Director, Information Systems
Provost and Vice President for Academic Affairs
Coppin State University
In the past, Coppin State University has successfully implemented many of the Blackboard Analytics modules. In spring 2012, Coppin started the Analytics for Learn implementation process, with a launch date of fall 2012. This presentation will share the details of our implementation effort, including strategy, feedback from faculty and students, and lessons learned. Got support from CIO and faculty champions buy in from faculty senate. Identify a few pilot courses. Training and webinars. Analytics Day, snacks faculty analytics session and training for deans. Reports for provosts deans and chairs dashboard features 10 standard reports, built with share point. Deans and chairs can see Across courses And departments, can drill down. Faculty see just their courses scatter plot course accesses against grades — does blackboard make a difference. Activity and grade may show at risk and achievers. Students can chart their own progress. Blackboard grade center side is critical. As so is grading consistently. Regular faculty use an timely responses to results. Course instruction method inconsistencies can skew results
Using Analytics to Assess the Impact of LMS Course Redesign Training John Fritz AVP, Instructional Technology & New Media University of Maryland, Baltimore County
Although we’ve long had an interest in learning analytics at UMBC, we’ve struggled to scale up development of our proprietary solutions. This session will show how we are using Blackboard Analytics for Learn (BA4L) to explore the LMS in much finer detail and assess the impact of faculty course redesign training. 95% students 75% faculty 60% courses. Only about 40 ba4l clients. Program offers 2500$ stipend. For training to redesign f2f for hybrid or online , teach, and evaluate. Folder depth ratio average items per folder divided by average folder depth. Too high or too low adversely affects use ability. Time consuming to build a cohort on bl4l. Content, tools, and assessment major report areas.