5. Objectives
• The state of the art
• Direct engagement with key stakeholders
• A comprehensive policy framework
http://sheilaproject.eu/
6. Inclusive adoption process
Macfadyen, L., Dawson, S., Pardo, A., Gašević, D., (2014). The learning analytics imperative and the sociotechnical challenge: Policy for complex systems. Research & Practice in Assessment,
9(Winter 2014), 17-28.
7. Methodology
Literature
- Policy
- Adoption
Academic staff
- Survey
- Focus groups
Students
- Survey
- Focus groups
Senior managers
- Survey
- Interviews
Experts
- Group concept
mapping
Policy
framework
Institutional
policy/strategy
Other stakeh.
- Workshops
- Committees
8. Methodology
Literature
- Policy
- Adoption
Academic staff
- Survey
- Focus groups
Students
- Survey
- Focus groups
Senior managers
- Survey
- Interviews
Experts
- Group concept
mapping
Policy
framework
Institutional
policy/strategy
Other stakeh.
- Workshops
- Committees
9. Adoption challenge
Leadership for strategic
implementation & monitoring
Tsai, Y. S., & Gasevic, D. (2017). Learning analytics in higher education – challenges and policies: a review of eight learning analytics policies. In Proceedings of the
Seventh International Learning Analytics & Knowledge Conference (pp. 233-242).
10. Adoption challenge
Equal engagement with
different stakeholders
Tsai, Y. S., & Gasevic, D. (2017). Learning analytics in higher education – challenges and policies: a review of eight learning analytics policies. In Proceedings of the
Seventh International Learning Analytics & Knowledge Conference (pp. 233-242).
11. Adoption challenge
Training to cultivate data literacy
among primary stakeholders
Tsai, Y. S., & Gasevic, D. (2017). Learning analytics in higher education – challenges and policies: a review of eight learning analytics policies. In Proceedings of the
Seventh International Learning Analytics & Knowledge Conference (pp. 233-242).
12. Adoption challenge
Policies for learning analytics practice
Tsai, Y. S., & Gasevic, D. (2017). Learning analytics in higher education – challenges and policies: a review of eight learning analytics policies. In Proceedings of the
Seventh International Learning Analytics & Knowledge Conference (pp. 233-242).
13. Methodology
Literature
- Policy
- Adoption
Academic staff
- Survey
- Focus groups
Students
- Survey
- Focus groups
Senior managers
- Survey
- Interviews
Experts
- Group concept
mapping
Policy
framework
Institutional
policy/strategy
Other stakeh.
- Workshops
- Committees
14. What is the state of the art?
What are the drivers?
What are the challenges?
15. Survey
• 22 countries, 46 institutions
• November 2016
NO P LA NS
IN P RE P A RA TION
IMP LE ME NT ED 2 13
15
16
The adoption of LA
Institution-wide Small scale N/A
16. Interviews
• 16 countries, 51 HEIs, 64 interviews, 78 participants
• August 2016 - January 2017
N O P L A N S
I N P R E P A R A T I O N
I M P L E M E N T E D
9 7 5
12
18
The adoption of learning analytics (interviews)
Institution-wide Partial/ Pilots Data exploration/cleaning
17. Motivations to adopt learning analytics
• To improve student learning performance – 40 (87%)
• To improve student satisfaction – 33 (72%)
• To improve teaching excellence – 33 (72 %)
• To improve student retention– 26 (57 %)
• To explore what learning analytics can do for our
institution/ staff/ students – 25 (54 %)
46 institutions
18. Motivations to adopt learning analytics
• To improve student learning performance – 40 (87%)
• To improve student satisfaction – 33 (72%)
• To improve teaching excellence – 33 (72 %)
• To improve student retention– 26 (57 %)
• To explore what learning analytics can do for our
institution/ staff/ students – 25 (54 %)
46 institutions
19. Motivations to adopt learning analytics
• To improve student learning performance – 40 (87%)
• To improve student satisfaction – 33 (72%)
• To improve teaching excellence – 33 (72 %)
• To improve student retention– 26 (57 %)
• To explore what learning analytics can do for our
institution/ staff/ students – 25 (54 %)
46 institutions
21. “People are thinking about learning analytics as a way
to try and personalise education and enhance
education. And actually make our education more
inclusive both by understanding how different students
engage with different bits of educational processes, but
also about through developing curricula to make them
more flexible and inclusive as a standard.”
22. “I think what we would be looking at is how do we
evolve the way we teach to provide better learning
outcomes for the students, greater mastery of the
subject.”
23. “We’re trying to understand better the curriculum that
needs to be offered for the students in our region.
And…I think importantly how our pedagogical model
fits that and deliver the best experience for our
students.”
24. Barriers to the success of learning analytics
• Analytics expertise – 34 (76%)
• A data-driven culture at the institution – 30 (67%)
• Teaching staff/tutor buy-in – 29 (64%)
• The affordances of current learning analytics technology – 29 (64%)
26. Implications
• Interests were high but experiences were premature.
• There was strong motivation in increasing institutional performance
by improving teaching quality.
• Key barriers were around skills, institutional culture, technology,
ethics and privacy.
27. Methodology
Literature
- Policy
- Adoption
Academic staff
- Survey
- Focus groups
Students
- Survey
- Focus groups
Senior managers
- Survey
- Interviews
Experts
- Group concept
mapping
Policy
framework
Institutional
policy/strategy
Other stakeh.
- Workshops
- Committees
29. With regards to learning analytics …
… what do academic staff ideally expect to happen?
… what do academic staff predict to happen in
reality?
Goal of the survey
30. 4 academic institutions
University of Edinburgh Carlos III Madrid
n = 81 n = 26
Open Universiteit University of Tallinn
n = 54 n = 49
from spring to fall 2017
31. 16 items, some examples
The university will provide me with guidance on how to access LA
about my students
The LA service will show how a student’s learning progress compares to
their learning goals/the course objectives
The teaching staff will have an obligation to act if the analytics show
that a student is at-risk of failing, underperforming, or that they could
improve their learning
32.
33.
34. University of Edinburgh:
• Ideal: LA will collect and present data that is accurate (M = 5.91) Q9
• Predicted: Providing guidance to access LA about students (M = 5.05) Q1
Carlos III de Madrid:
• Ideal: LA presented in a format that is understandable and easy to read
(M = 6.31) Q11
• Predicted: LA will present students with a complete profile of their
learning across every course (M = 5.27) Q12
Highest expectation values
35. Highest expectation values
Open Universiteit Nederland:
• Ideal: LA will collect and present data that is accurate (M = 6.60) Q9
• Predicted: Able to access data about students’ progress in a course that I
am teaching (M = 5.17) Q4
University of Tallinn:
• Ideal: Able to access data about students’ progress in a course that I am
teaching (M = 6.04) Q4
• Predicted: Able to access data about students’ progress in a course that I
am teaching (M = 5.49) Q4
36.
37. Lowest expectation values
University of Edinburgh:
• Ideal: Teaching staff will have an obligation to act if students are found to be at-
risk of failing or under performing (M = 3.65) Q14
• Predicted: Teaching staff will be competent in incorporating analytics into the
feedback and support they provide to students (M = 3.49) Q13
Carlos III de Madrid:
• Ideal: Teaching staff will have an obligation to act if students are found to be at-
risk of failing or under performing (M = 4.42) Q14
• Predicted: Teaching staff will have an obligation to act if students are found to be
at-risk of failing or under performing (M = 3.77) Q14
38. Lowest expectation values
Open Universiteit Nederland:
• Ideal: Teaching staff will have an obligation to act if students are found to be at-
risk of failing or under performing (M = 4.44) Q14
• Predicted: Feedback from analytics will be used to promote students’ academic
and professional skill development for future employability (M = 3.24) Q15
University of Tallinn:
• Ideal: Teaching staff will have an obligation to act if students are found to be at-
risk of failing or under performing (M = 4.80) Q14
• Predicted: Q14 (M = 3.82)
40. Goal
To better understand the viewpoints of academic staff on:
• Learning analytics opportunities in the HEIs from the
perspective of students, teachers and programs;
• Concerns related with adapting of learning analytics;
• Needed steps to adopt learning analytics at the HEIs
41. Study participants
• University of Edinburgh: 5 focus groups, 18 teaching staff
• Universidad Carlos III de Madrid: 4 focus groups, 16
teaching staff
• Open Universiteit Nederland: 2 focus groups, 5 teaching
staff
• Tallinn University: 5 focus groups, 20 teaching staff
42. Results: Expectations & LA opportunities
STUDENT
LEVEL
TEACHER
LEVEL
PROGRAM
LEVEL
Take responsibility for their
learning and enhancing their
SRL- skills
Assess the degree of success to
prevent students from begin
worried or optimistic about
their performance
Method to identify student’s
weaknesses and know where
students are with their progress
Understand how students
engage with learning content
Improve of the design and
provision of learning materials,
courses, curriculum and support
to students
Understand how program is
working (strengths and
bottlenecks)
Improve educational quality
(e.g. content level)
46. Results: concerns – program level
• Interpretation of learning:
• Was the right data collected?
• Were the accurate algorithms developed ?
• Was an appropriate message given for the students?
• Connecting LA to real learning – is this meaningful picture of
learning what is happening in online environments?
47. What we should consider?
• LA should be just one component of many for collecting
feedback and enhancing decision-making
• Involve stakeholders:
• Academic staff to in developing and setting up of LA
• Pedagogy experts involved to ensure data makes sense to
improve learning
• Provide training, communication!
48. What we should consider?
•Design of the tools that are:
•Easy to use
•Providing visualizations of data
•Not requiring mathematical/statistical skills
•Not taking a lot of time
•Considering ethical and privacy aspects
49. Student Views
Pedro Manuel Moreno Marcos
Department of Telematics Engineering
Universidad Carlos III de Madrid
pemoreno@it.uc3m.es
http://sheilaproject.eu/
50. Methodology
Literature
- Policy
- Adoption
Academic staff
- Survey
- Focus groups
Students
- Survey
- Focus groups
Senior managers
- Survey
- Interviews
Experts
- Group concept
mapping
Policy
framework
Institutional
policy/strategy
Other stakeh.
- Workshops
- Committees
52. Background
• 12 Items Survey
• Two Subscales:
• Ethical and Privacy Expectations
• Service Expectations
• 6 Distributions:
• Edinburgh (N = 884)
• Liverpool (N = 191)
• Tallinn (N = 161)
• Madrid (N = 543)
• Netherlands (N = 1247)
• Blanchardstown (N = 237)
http://sheilaproject.eu/
53. Ideal Expectation Scale Predicted Expectation Scale
Alternative Purpose Consent to Collect Identifiable Data Keep Data Secure Third Party Alternative Purpose Consent to Collect Identifiable Data Keep Data Secure Third Party
1
2
3
4
5
6
7
Item
Average
Location
Blanchardstown
Edinburgh
Liverpool
Madrid
Open University of the Netherlands
Tallinn
Ethical and Privacy Expectations http://sheilaproject.eu/
54. Keep Data Secure – Predicted Expectation Scale
Blanchardstown
Edinburgh
Liverpool
Madrid
Open University of the Netherlands
Tallinn
Strongly Disagree Disagree Somewhat Disagree Neither Agree nor Disagree Somewhat Agree Agree Strongly Agree
Response
10
20
30
40
50
Percentage
http://sheilaproject.eu/
55. Blanchardstown
Edinburgh
Liverpool
Madrid
Open University of the Netherlands
Tallinn
Strongly Disagree Disagree Somewhat Disagree Neither Agree nor Disagree Somewhat Agree Agree Strongly Agree
Response
10
20
30
Percentage
Consent to Collect – Predicted Expectation Scale http://sheilaproject.eu/
56. Ideal Expectation Scale Predicted Expectation Scale
ObligationtoAct
IntegrateintoFeedback
SkillDevelopment
RegularlyUpdate
CompleteProfile
StudentDecisionMaking
CourseGoals
ObligationtoAct
IntegrateintoFeedback
SkillDevelopment
RegularlyUpdate
CompleteProfile
StudentDecisionMaking
CourseGoals
1
2
3
4
5
6
7
Average
Location
Blanchardstown
Edinburgh
Liverpool
Madrid
Open University of the Netherlands
Tallinn
Service Expectations http://sheilaproject.eu/
57. Blanchardstown
Edinburgh
Liverpool
Madrid
Open University of the Netherlands
Tallinn
Strongly Disagree Disagree Somewhat Disagree Neither Agree nor Disagree Somewhat Agree Agree Strongly Agree
Response
10
20
30
Percentage
Course Goals – Predicted Expectation Scale http://sheilaproject.eu/
58. Blanchardstown
Edinburgh
Liverpool
Madrid
Open University of the Netherlands
Tallinn
Strongly Disagree Disagree Somewhat Disagree Neither Agree nor Disagree Somewhat Agree Agree Strongly Agree
Response
10
20
Percentage
Obligation to Act – Predicted Expectation Scale http://sheilaproject.eu/
59. Summary
• Beliefs towards learning analytics are not consistent.
• Emphasis on data security and improving learning.
http://sheilaproject.eu/
61. Background
• 18 focus groups
• 4 partners’ institutions
• 74 students
• Interviews: Around 1h
http://sheilaproject.eu/
62. Interests and expectations
• Improve the quality of teaching
• Better student-teacher feedback
• Better academic resources and academic tools to improve learning
• Personalized support
• Recommendation of learning resources
• Feedback from a system, via a dashboard
• Provide an overview of the tasks to be done in a semester → improve
curriculum design
http://sheilaproject.eu/
63. Awareness
• Students do not know what LA is, but they recognise its importance if
it can solve students’ problems
• Students are not generally aware of the data collected → Transparency
• Students have not checked the conditions they have accepted about
data
http://sheilaproject.eu/
67. • innovations in way network is delivered
• (investigate) corporate/structural alignment
• assist in the development of non-traditional partnerships (Rehab with the
Medicine Community)
• expand investigation and knowledge of PSN'S/PSO's
• continue STHCS sponsored forums on public health issues (medicine
managed care forum)
• inventory assets of all participating agencies (providers, Venn Diagrams)
• access additional funds for telemedicine expansion
• better utilization of current technological bridge
• continued support by STHCS to member facilities
• expand and encourage utilization of interface programs to strengthen the
viability and to improve the health care delivery system (ie teleconference)
• discussion with CCHN
Work
quickly and
effectively
under
pressure
49
Organize the
work when
directions are
not specific.
39
Decide how to
manage
multiple tasks.
20 Manage resources effectively.
4
2. Sort
3. Rate
1. Brainstorm
Group Concept Mapping
68. Onderwerp via >Beeld >Koptekst en voettekst Pagina 68
27 March 2014@HDrachsler 68 / 31
An essential feature of a higher education institution’s
learning analytics policy should be …
Group Concept Mapping
77. Cluster Map
1. privacy & transparency
2. roles & responsibilities
(of all stakeholders)
3. objectives of LA
(learner and teacher support)
4. risks & challenges
5. data management
6. research & data analysis
78. Rating Map – Importance
1. privacy & transparency
2. roles & responsibilities
(of all stakeholders)
3. objectives of LA
(learner and teacher support)
4. risks & challenges
5. data management
6. research & data analysis
Cluster Legend
Layer Value
1 5.08 to 5.27
2 5.27 to 5.46
3 5.46 to 5.65
4 5.65 to 5.84
5 5.84 to 6.03
79. Rating Map – Ease
1. privacy & transparency
2. roles & responsibilities
(of all stakeholders)
3. objectives of LA
(learner and teacher support)
4. risks & challenges
5. data management
6. research & data analysis
Cluster Legend
Layer Value
1 3.79 to 4.12
2 4.12 to 4.45
3 4.45 to 4.78
4 4.78 to 5.11
5 5.11 to 5.44
80. Rating Ladder Graph
importance ease
privacy & transparency
privacy & transparency
risks & challenges
risks & challenges
roles & responsibilities (of all stakeholders)
roles & responsibilities (of all stakeholders)
objectives of LA (learner and teacher support)
objectives of LA (learner and teacher support)
data management
data management
research & data analysis
research & data analysis
3.79 3.79
6.03 6.03
r = 0.66
81. Go Zone – Roles & Responsibilities
5
38
62
11
19
22
33
39 48
70
91
25
28
37
40
55
61
66
27
47 49
6.08
4.72
3.12
ease
3.83 5.48 6.59
importance
r = 0.26
55. being clear about the purpose of learning analytics
61. a clear articulation of responsibilities when it comes to the use of institutional data
82. Yi-Shan Tsai, Pedro Manuel
Moreno-Marcos, Ioana Jivet,
Maren Scheffel, Kairit Tammets,
Kaire Kollom, and Dragan
Gašević. (to appear). The
SHEILA framework: Informing
institutional strategies and
policy processes of learning
analytics. Journal of Learning
Analyitcs.
83. Methodology
Literature
- Policy
- Adoption
Academic staff
- Survey
- Focus groups
Students
- Survey
- Focus groups
Senior managers
- Survey
- Interviews
Experts
- Group concept
mapping
Policy
framework
Institutional
policy/strategy
Other stakeh.
- Workshops
- Committees
86. Methodology
Literature
- Policy
- Adoption
Academic staff
- Survey
- Focus groups
Students
- Survey
- Focus groups
Senior managers
- Survey
- Interviews
Experts
- Group concept
mapping
Policy
framework
Institutional
policy/strategy
Other stakeh.
- Workshops
- Committees
Hinweis der Redaktion
With senior managers, we were
EADTU (European Association of Distance Teaching Universities)
EUA (European University Association)
HeLF (Heads of e-Learning Forum)
EUNIS (European University Information Systems)
SNOLA (Spanish Network of Learning Analytics)
eMadrid
A survey question (multiple choices) provided 11 options for motivations specific to learning and teaching.
All related to institutional performance: league ranking, satisfaction survey, teaching excellence framework
But also dependent on teaching quality
Early stage - exploration
Most institutions seem to have incorporated all levels of goals into their planning or implementation of LA
Enhance self-regulation skills: provide data-based information to guide students
Improve learning support: curriculum, feedback, personalized support, pastoral care, timely support
Increase institutional performance: retention rate, student satisfaction, league ranking
13 options
Moderately-sized, large, critical
Three most mentioned issues regarding ethics and privacy
Interest is strong
Institutions were exploring what LA can do
Using LA to enhance teaching so as to increase institutional performance is the biggest motivation among managers
Barriers – skills, culture, technology, ethics and privacy
Qualitative data to get addition to LA data students’ perceptions and understandings about teaching and learning processes
Staff was worried that profiling of the students as e.g. low-performing might end up with the lost of motivation and anxiety.
Staff was wondering: shall I be objective?
How can I be objective
Hi – I’m Alex Wainwright… and I’m going to give an overview of the student survey results… this is going to cover response rates and some general insights obtained…
The student survey is composed of 12 items… and responses are made on two scales that correspond to a desired service… and what students expect in reality… so they reflect two levels of expectation…
Through the development and validation process we have identified two subscales… these refers to ethical and privacy expectations… such as whether students expect to provide consent for the collection of their educational data…
And the other subscales refers to service expectations… so this covers things such as whether students expect to receive updates on how their learning progress compares to a set goal….
As you can see… we have distributed the instrument at six different higher education institutions… with the highest response rate being at the open university of the netherlands…
All distributions have shown the scales to be valid and to also show excellent measurement quality….
Firstly… I am going to go over the ethical and privacy expectation items…
On this figure you can see the average responses to these items by expectation scale and location…. The x axis provides an indication of what the items refer to…
So we have beliefs about providing consent when data is used for an alternative purpose… or whether consent should be sought before distributing data to third party companies
What can be seen is that students ideal expectations are generally higher than predicted expectations – this is anticipated as it is a desired level of service…
Across both scales… however… we can see that the expectation that all collected data remain secure receives the highest average response… whereas… the expectation to provide consent before educational data is collected and analysed receives the lowest average response across these five items… and whilst students agree with this belief… it verges on indifference on the predicted expectation scale for the Spanish student sample….
It may be that students are open to universities collecting and analysing educational data… particularly as it is used for attendance purposes, for example…
Whereas… they have stronger beliefs toward universities abiding by data handling policies that will ensure that all data remains secure…
We can also look at these two particular ethical and privacy expectation items in more detail…
This figure shows the percentage of students responding in a certain way to the data security expectation… with darker colours reflecting a higher percentage of students responding that way…
And what is show is that… between 60 to 80% of students across all universities either agreed or strongly agreed with the expectation that universities will ensure data is kept secure…
For the consent to collect expectation… this figure shows that there is more variation in the responses…
For those students from Edinburgh, Liverpool, the Netherlands, and Blanchardstown… the largest response of around 30% is for strongly agree to this belief…
Whereas… the largest percentage of responses for Madrid and Tallinn… which was around 25%... Was for somewhat agree…
Looking at the service expectation items… we can that the average responses tend to be similar across locations…
Of particular note… the obligation to act is the item with the lowest response on average… with students in Madrid, the Netherlands, and Tallinn generally showing indifference to this belief on the predicted expectation scale...
The higher average responses… on the other hand… seem to be around aspects of self-regulated learning such as students expecting to receive a complete profile of their learning…. Making their own decisions on the analytics that they receive… and knowing how their progress compares to a set learning goal….
Looking into what are the highest and lowest average response items… we can also understand differences within each sample…
For knowing how progress compares to a set learning goal… between 20 to 35% of students across each sample agreed with this expectation… with around 4% disagreeing….
As for the obligation to act… the highest response rates are variable…
Around 20% of students in the Tallinn and Madrid samples somewhat disagreed with this expectation… For the Dutch students 24% expressed indifference to this belief… whereas in Liverpool and Blanchardstown around 28% showed agreement…
The output from the student survey shows that the expectations of students towards learning analytics are not consistent across each sample… with students generally showing variations in what they want from such services…
On the other hand… we can generally see that students expect a learning analytics service that emphasises data security… and provides tools that support learning as opposed to those that emphasise early interventions