Data driven blended learning: going from a heterogeneous classroom to personalized learning.
Presentation from 'In Focus: Learner analytics and big data', a CDE technology symposium held at Senate House on 10 December 2013. Conducted by Ernest Lyubchik (Founder, Head of Data and Algorithm Development, Selflab.com) and Dr Sara Hershkovitz (Head of Mathematics Dept, Center for Educational Technology, Israel).
Audio of the session and more details can be found at www.cde.london.ac.uk.
2. Agenda
• Selflab & CET are conducting an adaptive
learning experiment.
• Goal: validate Selflab’s adaptive technology.
• First stage introduced Selflab’s analytics.
• Presenting new instruction methods that
became possible using the analytics.
• First stage results.
• Future research.
3. Selflab
A paradigm shift in adaptive learning.
• Selflab's software-as-a-service platform
automatically serves any educational content
based on individual users' unique learning
needs.
• Selflab’s easy integration gives publishers
extensive data and analytics about their
content and their students.
5. The Center for Educational Technology (CET)
Goals
Promote
achievement and
academic excellence
in the 21st century
Create
equal opportunities
to all children
Activities
In its 42 years of activity, CET, the
leading educational NGO in
Israel, has established its expertise
and reputation by:
• Developing state-of-the-art printed and
digital content
• Developing interactive learning
objects, virtual labs and simulations in all
subjects
• Paving new ways in online and blended
learning programs for students and
teachers (PD)
• Creating online systems and tools for
assessment and evaluation
7. Mathematical Proficiency
Kilpatrick J., Swafford J. & Findell B. (2001): Adding It Up: Helping children learn mathematics.
Washington, DC: National Academy Press.
Common core state standards for mathematics, 2012
11. The aim is:
To focus on students’ readiness to learn, scaffolding
and building on his current knowledge and move
the student along an efficient trajectory according
to their own skills abilities and knowledge to the
next stage.
12. Pierce R., Stacey K. (2010): “Mapping Pedagogical Opportunities Provided by Mathematics analysis
Software” International journal of computers for Mathematical Learning, 15(1), pp. 1-20
13. Analytics – the first step for adaptive
• Current classroom instruction is limited by the
teachers knowledge of the students’ abilities.
• To fit the needs of the student, first we must
know what the student needs.
• A student making a multiplication mistake in
an Algebra question, needs help with
multiplication, not Algebra!
14. What are the students solving?
Hello Ernest [exit]
Difficulty scale
Laboratory
Write < or > or =
Check my answer
Back
15. The shape is a single unit.
Write a fraction and a mixed number that match the illustration
Fraction
Check my answer
Mixed number
16. Pizzas at the pizza stand are divided to 3 equal slices. Make an order for 6 slices of pizza
Check my answer
18. Experiment
• 6 classes of 4th grade, from 2 different
schools, were selected for the first phase.
• The students began learning fractions without
prior instruction, receiving computerised
instruction and exercises.
• All data was monitored,
and the teachers received
reports, while the system
itself did not activate
adaptive personalisation.
19. Reporting and Analytics
• CET was provided with content reports, and
the content on the platform was supervised.
• The teachers received both custom requested
reports of the student’s performance, and a
real time tracking system to oversee the
progress of the students.
22. New Instructional Methods
• Students receive instruction from the
program. However, some students still require
instruction from the teacher.
• Instead of presenting the class with frontal
instruction, the teacher can work directly with
those who require assistance, helping them
with targeted sessions.
23.
24.
25.
26.
27. Feedback
• Students’ feedback:
“This is right where I am now, how did you
know I had a problem with that?”
• Teachers’ feedback:
“The class is more organized”
“I know what they really understood”
“The students love the program”
“Can we allow sixth graders to use it?”
28. Big impact for low achievers
• The teachers pushed for adding sixth graders who
had problems with the material to the
experiment.
• The sixth graders who joined were low
achievers, and unmotivated students. In just a
few lessons, the students reported increased
motivation, confidence and interest.
• The previously low achievers improved their
performance, they achieved similar ability grades
as the rest.
30. Results
• Personalised learning students made on average
%14 mistakes compared to 25% of frontal
instruction groups. 44% less mistakes.
• The measured performance gain of the
personalised learning students was found
significant by T-Test, ALPHA=0.01.
• Personalised learning led to higher levels of
motivation and lower levels of class disruptions
and stress. (precise data still being gathered).
31. Future plans
• Going adaptive – individual learning pathways.
• Understanding student’s abilities rather than
student’s score.
• Data driven analysis of content.
But what is the difference
between success rate and ability?