The document discusses Desire2Learn's analytics and insights tools for education including Degree Compass, which uses predictive modeling to recommend courses and help students succeed. It outlines Desire2Learn's analytics capability maturity model and portfolio including Analytics Essentials, Insights reporting tools, and Degree Compass. Research data shows Degree Compass improves grades, course completion rates, and reduces achievement gaps.
1. You have data
We provide Insights™
Big Data in Education:
Theory and Practice
Michael Moore, MSCIS
Sr. Advisory Consultant – Analytics
2. ILP - Analytics Capability and
Maturity Model
Stage Four
Insight and Information Value
What do you want to happen
for you??
Stage Three
Stage One
Stage Two
What do I want to happen?
Advanced Adaptive
What will happen?
What has happened?
What is happening?
Data
Reporting
Access
Advanced Predictive
Predictive
Risk
Modeling
Forecasting
Strategic
Optimization
D2L Integrated Learning and Advanced Analytics Platform
4. Optimized for Potential Performance
“Our primary motivation for deploying Degree Compass
was to respond to the unique success and retention
needs of our complex student population.”
Dr. Tristan Denley, | Austin Peay State University
5. What ingredients go into the ratings?
Global centrality
Major centrality
Grade prediction
7. Research Data:
Grade Predictions
• Typical probability of
achieving A or B in a
course – 62%
•
Degree Compass™
recommended courses –
probability of A or B in a
course – 85%
9. Research Data:
Attainment Gap
Minimized
performance disparity
based on
•
•
•
•
Socioeconomic status
Ethnicity
Gender
First-generation
enrollments
“Typical achievement
gap of 20% is closed
to a mere 6%.”
http://www.desire2learn.com/learninginsights/2013/11/Looking-to-Improve-Success/
11. Degree Compass™ Logical Diagram
Student application such as
Ellucian’s Banner Student
Advisor application such as
Ellucian’s Degree Works or
CAPP
Degree Audit
System
SIS
High School
transcripts, GPA, SAT
s, ACTs, Class
Rank, etc.
Program Audit or Inventory Courses, Schedules, Core and
Elective Requirements, Courses
taken vs courses still needed for
degree completion, etc.
Degree
Compass™
D2L’s Personalized, course
recommendation app
12. Degree Compass™ Details
• SIS Systems
– Banner v8.0
• Portals
– Banner Self-Service v8.0
– Luminis Portal v4.0
• Degree Audit Systems
– Degree Works v4.1.0
– CAPP v8.5.3
• Web Browsers
– Chrome, IE, Firefox, Safari
• Mobile Platforms
– iOS v4.0 (native)
– Android v2.3 (non-native)
13. Science Behind Success
• Degree Compass™ composed of two predictive
modeling engines and a predictive algorithm
– Grade Prediction Engine/Model
• Provides accurate estimate of final grade student will likely receive
– Centrality Prediction Engine/Model
• Provides accurate ranking of a program’s courses at the institution
• Degree Compass™ uses several inputs to drive
predictive engines and algorithm
– Student Information System (SIS)
• Banner SIS
– Degree Audit Systems (DAS)
• DegreeWorks
• CAPP
14. Degree Compass™ Roadmap
• TODAY: Which {course} should I choose
Incoming Freshman focus
• Which {program} should I choose
• Which {career} should I choose
15. Mapping this to D2L’s Solution Set
• Degree Compass™ - what are your
predictors for success in the program
• Student Success System - at what points
should we be concerned
• These two together are a powerful analog
for the onboarding
16. Adaptive Learning
• Knowillage LeaP
• Adaptive learning engine
• Personalized learning experience
What if textbooks could learn . . . from you?
25. Get to Know Your Big Data
Let the dataset change your mindset.
Subtitle
www.Desire2Learn.com
26. Questions to Think About…
What if…
• you had a reporting tool to directly support
institutional improvement initiatives?
• you had a reporting tool which was focused on
support of student retention and improvement?
• students could receive personalized content
before they knew what they needed?
• you could identify trends and make decisions
based on free text in discussion forums?
27. Questions?
Michael Moore, MSCIS
Sr. Advisory Consultant - Analytics
Direct 888.772.0325 x6604
Twitter: @MikeMooreD2L
Michael.Moore@Desire2Learn.com
Thank You
Subtitle
www.Desire2Learn.com
28. Let’s transform
teaching and learning,
together.
Desire2Learn, Campus Life, CaptureCast, Desire2Learn Binder, myDesire2Learn, Insert Stuff, Insert Stuff Framework,
Instructional Design Wizard, and the molecule logo are trademarks of Desire2Learn Incorporated.
Subtitle
The Desire2Learn family of companies includes Desire2Learn Incorporated, D2L Ltd., Desire2Learn Australia Pty Ltd,
Desire2Learn UK Ltd, Desire2Learn Singapore Pte. Ltd. and D2L Brasil Soluções de Tecnologia para Educação Ltda.
www.Desire2Learn.com
Notes de l'éditeur
How did we do that? Where did we start?First generation LMS was first step way back in 1999 – Stage OneFirst generation analytics technology added - Stage TwoSecond generation analytics technology with deeper/richer learning data curation (optimization of Stage Two tools and technology) – Stage ThreeSecond generation predictive and personalized/adaptive learning analytics with the Learner in absolute control of their destiny. Institutions are beacons for learners and drive focus and guidance with predictive and adaptive tools and technology. Institutions who employ these types of tools and technology will attract the largest student population and deliver the most skilled and capable graduates into the field. – Stage FourStage Four – I made this blue like Stage One as this will become the new normal or new baseline for learning in the 21st century. It will be the benchmark/baseline for all learning tools and technology moving forward. See next slides for more details on what is possible for learning as we move towards the end of the second decade of the 21st century. D2L is building the foundation (ie APIs, analytics/predictive analytics, adaptive, gaming, etc.) for Stage 4 entrance and expansion.Data access – just dataReporting/OLAP - what happenedForecasting – why did it happenPredictive modeling – what will happenOptimization (real-time predictive analysis)– what is the best that could happenSense and respond. Predict and Act.Marketing blurb:Already using the learning environment data available to report on key learning outcomes, student engagement and enrolment metrics as well as student grades data, Desire2Learn’s vision for big data in education was to move the institution from traditional activity reporting functions to a big data-driven framework with learning and academic analytics functionality at its core. By developing partnerships with key industry leaders in enterprise analytics applications, the Desire2Learn Analytics portfolio for education would be transformational.Understanding that big data concepts were new to education, Desire2Learn Analytics was specifically packaged into bundled offerings not only to suit different institutional reporting needs but also to address different institutional strategies around big data. Further, by developing a strategic roadmap to include predictive modules in their offering, Desire2Learn’s Analytics portfolio would offer an unprecedented suite of products with the capability to tap into the vast amounts of big data available in education today. Desire2Learn Analytics Portfolio delivers a multi-tiered analytics solution that offers customers a path forward to manage the analytics initiatives that are critical to their institutional effectiveness. Whether those analytics initiatives focus onincreasing operational efficiency, optimizing learning outcomes or creating the conditions for learner success, Desire2Learn’s Analytics Portfolio delivers two analytics solutions that are integral to your institution's strategic process improvement efforts.
Compatibility: Standalone componentWeb-based app (desktop or mobile)Supports IE v8.0 and above, Chrome, FirefoxIntegrates with SIS:BannerWorking on PeopleSoft integrationIntegrate with DAS:Degree Works CAPP
Total = % who answered correctlyUpper 27% = the percentage of ppl in the top quartile who answered the question correctlyDiscrimination index = shows discrimination between upper and lower quartile performers (generally the higher the better)If there is a negative, you prob want to remove that qu`estion from the examReliability coefficient:Only works for tests that are designed to test a coherent body of knowledge (so questions are correlated)High reliability indicates that the test is a reliable/good gauge of the student’s knowledge in a specific area (b/c noticeable patterns (correlation) will occur in student marks on the questions)B/w 0 and 1 -> >75% is very reliableHigher is goodPoint biserialCorrelation coeffecientCorrelation b/w getting the question correct and doing well on the overall examHigher is betterNegative is problematicResponse frequenciesGood at indicating effectiveness of distractor questions
All about aggregating outcomes across levels within the org
Shows relationships between learning outcomes and other learning outcomes at various organizational levelsInstitutional levelProgram levelDepartment levelCourse template level
Let the dataset change your mindset.We need to change our mindset when thinking about transforming our schools, and we need to be willing to "thrive on the unknown, appreciate ambiguity, and relish being different," to be willing to implement "yet-to-be-proven ideas," and to "focus on being different first and then on being better"--all of which take courage and an ability to learn as you move forward.