Learning Analytics - what is it and why is it difficult (InfiniteConf Bytes)
Download as PPTX, PDF4 likes666 views
The document discusses learning analytics, which refers to the measurement, collection, analysis and reporting of student data to understand and optimize learning. Some applications of learning analytics include early alert systems to identify at-risk students, course recommendations, adaptive learning, and curriculum design. However, learning analytics also faces challenges because it is difficult to determine if it actually improves student outcomes, how to address privacy and ethics concerns, and how to ensure buy-in from end users.
![So what’s the problem?
By Asit K. Ghosh Thaumaturgist (Own work) [CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia
Commons](https://image.slidesharecdn.com/learninganalytics-infiniteconfbytes-170508145337/85/Learning-Analytics-what-is-it-and-why-is-it-difficult-InfiniteConf-Bytes-14-320.jpg)
Learning Analytics - what is it and why is it difficult (InfiniteConf Bytes)
- 1. Learning Analytics – what is it and why is it so difficult? Samantha Ahern – Digital Education, ISD, University College London
- 2. The author
- 3. Overview What is it? What are the applications? Where does the data come from? Why is it difficult? Conclusions
- 5. What is it? Learning analytics is the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs. Siemens & Gašević, 2012 Guest Editorial - Learning and Knowledge Analytics. Educational Technology & Society, 15 (3), 1–2. Siemens, G., & Gašević, D. (2012)
- 6. What are the applications? Early Alert and Student Success Course Recommendation Adaptive Learning Curriculum Design
- 7. Early Alert & Success At present, ~20 million university students in the EU ~7million will never complete their degree Certain sub-groups are over-represented E.g. mature age students, BME students Universities vary in their retention rate E.g. some record drop-out rates as high as 43% Statistics from University of Exeter
- 8. Early Alert & Student Success cont’d https://youtu.be/-BI9E7qP9jA
- 9. Early Alert & Student Success cont’d https://flic.kr/p/95X3Ru
- 13. Curriculum Design cont’d Pg 67, Learning Analytics Explained, Niall Sclater, 2017
- 14. So what’s the problem? By Asit K. Ghosh Thaumaturgist (Own work) [CC BY-SA 3.0 (http://creativecommons.org/licenses/by-sa/3.0) or GFDL (http://www.gnu.org/copyleft/fdl.html)], via Wikimedia Commons
- 15. Does it work? Nick Youngson - link to - http://nyphotographic.com/ Evidence.laceproject.eu/evidence/country-map
- 17. We have it, should we use it? https://analytics.jiscinvolve.org/wp/files/2017/02/2017-02-22-Jisc-LAN-Boulton-et-al.pdf
- 18. What about the end users? http://www.animatedheroes.com/gromit
- 19. Is that an elephant? https://commons.wikimedia.org/wiki/File:Elephant%27s_tea_party,_Robur_Tea_Room,_24_March_1939,_by_Sam_Hoo d_(8739115901).jpg https://de.slideshare.net/Drachsler/delicate-checklist-to-establish-trusted-learning-analytics
- 21. Further Reading • JISC Learning Analytics: https://www.jisc.ac.uk/reports/learning-analytics-in-higher- education • Open University: http://www.open.ac.uk/iet/main/research-innovation/learning-analytics • S Ahern’s Digi Ed blog posts: http://blogs.ucl.ac.uk/digital-education/category/our- views/sams-scribbles/ • The Royal Society – Machine Learning: https://royalsociety.org/topics- policy/projects/machine-learning/ • Royal Statistical Society – Opportunities and Ethics of Big Data: http://www.rss.org.uk/Images/PDF/influencing-change/2016/rss-report-opps-and-ethics- of-big-data-feb-2016.pdf
- 22. Questions
Editor's Notes
- #5: End clip at 3:10
- #9: Large focus on engagement metrics – retention a key focus Predictive models built using historic data,
- #10: What does success look like – varies by individual and discipline HEFCE studies on learning gain – is this any better? Engagement often used – no clear agreement on what it is, beware of the low hanging fruit
- #11: Content-based filtering &/ collaborative filtering Possibly demographic-based, increased risk Systems tend to use historical data of past student performance plus student grade history and demographics
- #12: Horizon Report 2016 definition: Adaptive learning technologies apply learning analytics through software and online platforms, adjusting to individual students’ needs Commonly part of a blended learning design. Predominantly used in the sciences, rule based systems Development boosted by Bill & Melinda Gates Foundation – Next Gen Courseware Challenge (US $20Million)
- #16: Exeter systematic review of LA interventions, lack of empircal testing and evaluation – only identified 4 studies that evaluated the effectiveness of LA interventions, 20 studies reviewed Main issue was weak design LACE project evidence base: evidence.laceproject.eu
- #17: Drown the bunnies - Simon Newman, president of Mount St. Mary’s University, in Maryland, told faculty members: “This is hard for you because you think of the students as cuddly bunnies, but you can’t. You just have to drown the bunnies … put a Glock to their heads.”
- #18: University of Exeter data sources, green = curently in use, amber = planned, red = not planned How much do these sources tell us about student learning and how much about the system? Are they true proxies or just low hanging fruit Should we include all this data, is some of it truly relevant. What about privacy?
- #19: Are tutors / academics equipped or willing to act on the information provided A lot of discussion on making analytics safe for students - ensuring correct interpretation
- #20: GDPR – profiling, informed consent, automated decision making - https://ico.org.uk/for-organisations/data-protection-reform/overview-of-the-gdpr/individuals-rights/rights-related-to-automated-decision-making-and-profiling/ Duty of care? – is there an obligation to act – legally no, morally? What about human rights and privacy? Delicate framework: https://de.slideshare.net/Drachsler/delicate-checklist-to-establish-trusted-learning-analytics