Student consent in learning analytics: Finding a way out of the labyrinth

ELESIG ‘Conversations’ Webinar 29 January 2018
11:00 – 12:00 GMT
Student consent in learning analytics:
Finding a way out of the labyrinth
Image credit: https://pixabay.com/en/maze-graphic-render-labyrinth-2264/
By Paul Prinsloo
(University of South Africa)
@14prinsp

11:00 – 12:00 GMT
Acknowledgements
• Since 2013 my thinking about ethical considerations in the
collection, analysis and use of student data has been shaped by,
inter alia, my collaboration with Dr Sharon Slade (Open
University). I am deeply indebted to her for her input and her
inspiration
• I furthermore don’t own the copyright of any of the images in
this presentation. I hereby acknowledge the original copyright
and licensing regime of every image and reference used.
• This work (excluding the images) is licensed under a Creative
Commons Attribution 4.0 International License.

11:00 – 12:00 GMT
Image credit: https://www.quora.com/Whats-the-simplest-way-to-create-an-SMS-opt-in-marketing-campaign
What is the scope and intention of students’ consent when they sign
their learning contracts at registration? Do they provide permission
that we may…
• combine their demographic and learning behavior data to assess their potential
for non-payment, failure or success
• inform their teaching staff and course support teams accordingly
• allocate or withdraw resources according to the analysis of their data
• determine their future enrollment prospects based on their past behavior
• personalise/individualise their curricula, assessment, pedagogy and time
allowed to complete the course
• use their behavioral data and performance in one course to determine the
possibility of their success in another course
• publish the (anonymised) findings?

11:00 – 12:00 GMT
Image credit: http://notbuyinganything.blogspot.co.za/2012/01/opting-out-increasingly.html
Will they be allowed to opt out of
having their personal identifiable
data collected, analysed and used
when the collection and analysis will
affect their choices, access to
resources and standing (whether
positively or negatively)?
And if allowed, what are the implications for
• them – do they understand the risks/value?
• the teaching and support team
• the institution – our fiduciary and quality assurance duty?
• our (and their) understanding the complexities of
teaching and learning?

11:00 – 12:00 GMT
Image credit: http://chrisschweppe.com/?p=340Image credit: https://www.quora.com/Whats-the-simplest-
way-to-create-an-SMS-opt-in-marketing-campaign
Thinking in binary terms does
not really help. We have to
find a more nuanced
understanding of student
consent and student
participation in making sense
of their learning journeys

Imagecredit:https://www.flickr.com/photos/haydnseek/2534088367
Towards a (more) nuanced understanding of
informed consent in learning analytics
Image credit: http://www.yourtango.com/201168184/facebook-relationship-status-what-does-its-
complicated-mean
11:00 – 12:00 GMT

11:00 – 12:00 GMT
How do we talk about student consent in
the collection, analysis and use of their
data considering…

11:00 – 12:00 GMT
Knowing more about them may alert us to students who
experience distress, or where their silences/absences in
online learning environments may allow us to reach out
and make a difference. If only we knew…
Knowing more about them may allow us to allocate
resources more effectively
Knowing more about them and the way they engage, or
choose not to engage, may enrich our (and their)
understanding of the complexities of learning and
challenge some of our/their beliefs and assumptions

11:00 – 12:00 GMT
The asymmetries in power
between students and the
providing institutions - where
students do not know what
data are collected, when the
data are collected and used
by whom and for what
purpose, how the data are
governed and how this may
affect the current and future
registrations
Image credit: http://faithandheritage.com/2012/02/the-fifth-commandment-versus-egalitarianism/

11:00 – 12:00 GMT
That’s no Learning Management System. That’s a layered
system of surveillance, tracking what they do and don’t do,
when they access (which) resources, how many times they
access these resources, how and when they participate (if),
how many videos they watch (till the end), who they interact
with (and who not) and then combine that data with their
demographic and historical learning and behavioral data to
classify them into categories of ‘students-like-you’ and then
based on this, shape their learning journeys, access to
resources and future registrations.

Source credit: http://2017trends.hackeducation.com/data.html
11:00 – 12:00 GMT

“Students can [should be able to] see how these systems
work, you know – the decisions that are human-made
and the decisions that are machine-made and the
decisions that are historical and the decisions that are
structural. They worry that they are being set up to
fail. They worry that their data – their very identities –
are being weaponized against them. It’s not simply “the
algorithm” that causes educational inequalities to
persist. Students know that. Algorithms are just
becoming an easier way to justify unjust decision-
making”
11:00 – 12:00 GMT
Source credit: http://2017trends.hackeducation.com/data.html

Our data are not something
separate from our identities,
our histories, our beings.
Our data are an integral,
albeit informational part of
our being. Data are
therefore not something we
own and can give away. We
don’t own our data but we
are, increasingly, constituted
by our data.
See Floridi, L. (2005). The ontological
interpretation of informational privacy.
Ethics and Information Technology, 7(4),
185-200.
Image created from https://pixabay.com/en/steampunk-man-male-person-
fantasy-1809590/ and https://pixabay.com/en/matrix-network-data-
exchange-1013611/
11:00 – 12:00 GMT

11:00 – 12:00 GMT
A more nuanced understanding of
student consent needs to consider…
• The different sources of data
• The increasing automation of the collection, analysis and use
of (student) data
• The scope and intention of ‘consent’ and how it may change
in different types of analytics
• The brokenness of our data

Three sources of data
Directed
A digital form of
surveillance
wherein the
“gaze of the
technology is
focused on a
person or place
by a human
operator”
Automated
Generated as “an
inherent,
automatic function
of the device or
system and
include traces …”
Volunteered
“gifted by users
and include
interactions
across social
media and the
crowdsourcing of
data wherein
users generate
data” (emphasis
added)
Kitchen, R. (2013). Big data and human geography: opportunities, challenges and risks. Dialogues in Human
Geography, 3, 262-267. SOI: 10.1177/2043820613513388
11:00 – 12:00 GMT

(1)
Humans
perform the
task
(2)
Task is shared
with
algorithms
(3)
Algorithms
perform task:
human supervision
(4)
Algorithms
perform task: no
human input
Seeing Yes or No? Yes or No? Yes or No? Yes or No?
Processing Yes or No? Yes or No? Yes or No? Yes or No?
Acting Yes or No? Yes or No? Yes or No? Yes or No?
Learning Yes or No? Yes or No? Yes or No? Yes or No?
Danaher, J. (2015). How might algorithms rule our lives? Mapping the logical space of algocracy. [Web log post]. Retrieved from
http://philosophicaldisquisitions.blogspot.com/2015/06/how-might-algorithms-rule-our-lives.html
Consent in the context of human-algorithm
interaction
11:00 – 12:00 GMT

Source credit: http://timoelliott.com/blog/2013/02/gartnerbi-emea-2013-part-1-analytics-moves-to-the-core.html
11:00 – 12:00 GMT
What do students consent to?

Citation: Pink, S., Ruckenstein, M., Willim, R., & Duque, M. (2018). Broken data: Conceptualising data in an
emerging world. Big Data & Society, 5(1), 2053951717753228.
11:00 – 12:00 GMT

Page credit: http://er.educause.edu/articles/2012/7/learning-analytics-the-new-black
2012
11:00 – 12:00 GMT

Page credit: https://library.educause.edu/resources/2015/10/the-predictive-learning-analytics-revolution-leveraging-learning-data-for-student-
success
2015
11:00 – 12:00 GMT

Page credit: https://icde.memberclicks.net/assets/RESOURCES/anne_la_report%20cc%20licence.pdf
2016
Page credit: https://icde.memberclicks.net/assets/RESOURCES/anne_la_report%20cc%20licence.pdf
11:00 – 12:00 GMT

11:00 – 12:00 GMT
Unthinkable, and totally off the radar was any
consideration of the notion of student choice to, at a
minimum, to provide informed consent for the collection,
analysis and use of their data, or at most, participate as
equals in the sense-making of their learning journeys

11:00 – 12:00 GMT
Slade, S., & Prinsloo, P. (2013). Learning analytics: Ethical issues and dilemmas. American Behavioral Scientist, 57(10), 1510-1529.
2013

11:00 – 12:00 GMT
Source credit: https://www.open.ac.uk/students/charter/sites/www.open.ac.uk.students.charter/files/files/ecms/web-content/ethical-
use-of-student-data-policy.pdf
2014

11:00 – 12:00 GMT
Opting out is not
an option
Image credit: http://notbuyinganything.blogspot.co.za/2012/01/opting-out-increasingly.html

Page credit: http://www.npr.org/sections/ed/2017/01/11/506361845/the-higher-ed-learning-revolution-tracking-each-
students-every-move
2017
11:00 – 12:00 GMT

Imagecredit:https://pixabay.com/en/binary-code-man-display-dummy-face-1327512/
Page credit: https://www.edsurge.com/news/2017-06-13-from-high-school-to-harvard-students-urge-for-clarity-on-privacy-
rights?utm_content=buffer8dd71&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer
3 June 2017
11:00 – 12:00 GMT

Source link: https://twitter.com/SwedishCanary/status/878622084141690881
11:00 – 12:00 GMT

11:00 – 12:00 GMT
“Ideally, we should get rid of learning analytics
altogether. It is a colonialist, slave-owning,
corporatizing, capitalist practice that enacts
violence, yes violence, against the sanctity of a
learner’s privacy, body and mind” (Hathcock, 2018).
Source credit: https://aprilhathcock.wordpress.com/2018/01/24/learning-agency-not-analytics/

Imagecredit:https://pixabay.com/en/binary-code-man-display-dummy-face-1327512/
Page credit: https://www.amazon.com/Ethical-
Futures-Qualitative-Research-
International/dp/1598741411
Page credit:
https://press.anu.edu.au/publications/series/centre-
aboriginal-economic-policy-research-caepr/indigenous-data-
sovereignty

11:00 – 12:00 GMT
Griffiths, D. (2017, September). An Ethical Waiver for Learning Analytics?. In European Conference on Technology
Enhanced Learning (pp. 557-560). Springer, Cham.

11:00 – 12:00 GMT
Image credit: https://www.flickr.com/photos/jackskellington101/426791087
“… the ethical processes of academic research seem a
straitjacket which prevents them from their methods”
(Griffiths, 2017, pp. 2-3)
Griffiths, D. (2017, September). An Ethical Waiver for Learning Analytics?. In European Conference on Technology Enhanced Learning (pp.
557-560). Springer, Cham.
Learning analytics resemble more Operations Research
and as such…

11:00 – 12:00 GMT
Source credit: https://www.technologyreview.com/s/609132/dont-let-regulators-ruin-ai/
Vol. 120 | No. 6, p. 73

11:00 – 12:00 GMT
“Policymakers who wish to champion
growth should embrace a stance of
“permissionless innovation.” Humility,
collaboration, and voluntary solutions
should trump the outdated “command and
control” model of the last century. The age
of smart machines needs a new age of
smart policy.”
Source credit: https://www.technologyreview.com/s/609132/dont-let-regulators-ruin-ai/
Andrea O’Sullivan is a program manager with the Mercatus Center, a free-
market-oriented think tank at George Mason University’s Technology Policy
Program.

11:00 – 12:00 GMT
Higher education has always collected, analysed and used
student data – so what has changed?
Why is consent suddenly an issue?
Image credit: https://en.wikipedia.org/wiki/Scholasticism

11:00 – 12:00 GMT
IN THE PAST AT PRESENT
Data sources Demographic and learning data at
specific points in the learning
journey: data application,
registration, class registers,
assignments, summative
assessment, personal
communication
Continuous directed and
automated collection of data
from a range of data sources –
student administration, learning
management system (LMS),
sources outside of the LMS
Data use Reporting purposes, operational
planning on cohort, group level by
management, institutional
researchers
Descriptive, diagnostic, predictive
and prescriptive on group/cohort
level
Plus individualised, often real-
time use of data to inform
pedagogy, curriculum,
assessment, student support by
faculty, students and support
staff
Who used the
data
Management, institutional
researchers, planners, quality
assurance and HR departments
Plus researchers, faculty,
students and support staff

11:00 – 12:00 GMT
IN THE PAST AT PRESENT
Who did the
collection,
analysis and
who used the
data
Humans Increasingly humans in
combination with algorithmic
decision-making processes
Temporal aim Retrospective/historical data to
make predictions with regard to
budget, future enrollments &
resource allocation on institutional
level
Plus real-time data for real-time
interventions
Default Forgetting Remembering
Personal
identifiers
Anonymised, aggregated data Plus re-identifiable data
Personal/ised data
Oversight/
data
governance
Broad institutional oversight.
Ethical Review Board (ERB)
approval for research purposes
Approval, oversight and
governance highly complex and
contested

11:00 – 12:00 GMT
Why is it necessary to consider ‘consent’ in
the collection, analysis and use of student
data?
• We have access to greater volumes of data, an increased
velocity of data and great granularity of student data from a
variety of sources than ever before. With this increased scope
of access, the increased capacity of our hardware/software
and the dangers of epistemic arrogance
• The brokenness of our data
• The unintended consequences and the potential of harm
• The importance to move away from students as data objects
to students as equal partners, as owners of the data
• We have a fiduciary duty and moral obligation

11:00 – 12:00 GMT
Image credit: https://uvmbored.com/event/office-undergraduate-research-weekly-workshops/2017-12-07/
Central to considering student consent is the
question: Is learning analytics….?

Willis, J. E., Slade, S., & Prinsloo, P. (2016). Ethical oversight of student data in learning analytics: A typology derived from a cross-
continental, cross-institutional perspective. Educational Technology Research and Development, 64, 881-901. DOI: 10.1007/s11423-
016-9463-4 http://link.springer.com/article/10.1007/s11423-016-9463-4
Definition, oversight and accountability
An interpretative multiple-case study: Indiana University, Open University (UK) and
the University of South Africa (Unisa)
11:00 – 12:00 GMT

Typology: Learning
analytics as…
Approval/oversight/
accountability
Research Formal, well-defined processes
An emerging form of
research
Undefined, unclear
Our current processes do not allow for any oversight
Scholarship of teaching and
learning
Undefined, unclear
Consent normally not required. Oversight? Student
complaints, feedback
Dynamic, synchronous and
asynchronous sense-
making
Undefined, unclear
Automated Undefined, unclear
Participatory process and
collaborative sense-making
All stakeholders are involved – may need broad, blanket
consensus at the beginning of each course – oversight by
the highest academic decision making body. Important
here is the role of students as collaborators in sharing
interpretation, governance, quality assurance, integrity
of data

11:00 – 12:00 GMT
Griffiths, D. (2017, September). An Ethical Waiver for Learning Analytics?. In European Conference on Technology
Enhanced Learning (pp. 557-560). Springer, Cham.
Learning analytics has a closer resemblance to
Operations Research than traditional education research
and therefore …

11:00 – 12:00 GMT
Source credit: https://depts.washington.edu/cfar/sites/default/files/uploads/scientific-programs/health-systems/Mini2006OREthicsGloyd.pdf

Source credit: https://depts.washington.edu/cfar/sites/default/files/uploads/scientific-programs/health-systems/Mini2006OREthicsGloyd.pdf

11:00 – 12:00 GMT
Sclater, N. (2017, June 30). Consent and the GDPR: what approaches are universities taking? Retrieved from
https://analytics.jiscinvolve.org/wp/2017/06/30/consent-and-the-gdpr-what-approaches-are-universities-
taking/

11:00 – 12:00 GMT
CONSENT TYPE OF DATA
Not ask for consent Non-sensitive data as long as the data
and analysis can be considered as of
legitimate interest or public interest
Ask for consent Sensitive data (under the GDPR, this
will be called ‘special category data)
Ask for consent When the data and analytics will
directly link to interventions that will
affect the student
Sclater, N. (2017, June 30). Consent and the GDPR: what approaches are universities taking? Retrieved from
https://analytics.jiscinvolve.org/wp/2017/06/30/consent-and-the-gdpr-what-approaches-are-universities-
taking/

11:00 – 12:00 GMT
• Data are “framed technically, economically, ethically, temporally,
spatially and philosophically” and do not exist independently “of
the ideas, instruments, practices, contexts and knowledges used
to generate, process and analyse them” (Kitchin 2014, p. 2).
• Data are never “neutral, objective, and pre-analytic” (Kitchin 2014,
p. 2)
Source: Prinsloo, P., & Slade, S. (in press). Student consent in learning analytics: the devil in the details? In J.
Lester, C. Klein, H. Rangwala, and A. Johri (Eds), Learning analytics in higher education: Current innovations,
future potential, and practical applications. Routledge.
Under what circumstances do the following data become ‘sensitive’
resulting in potential bias, discrimination and exclusion?
• Gender
• Race
• Home address
• Occupation

11:00 – 12:00 GMT
Madden, M., Gilman, M., Levy, K., & Marwick, A. (2017). Privacy, Poverty, and Big Data: A Matrix of
Vulnerabilities for Poor Americans. Wash. UL Rev., 95, 53.

11:00 – 12:00 GMT
Source credit: https://newrepublic.com/article/146710/injustice-algorithms

11:00 – 12:00 GMT
Source credit: https://www.commonsense.org/education/privacy/blog/digital-redlining-access-privacy

11:00 – 12:00 GMT
Towards a (more) nuanced understanding
of informed consent in learning analytics
Image credit: https://pixabay.com/en/nuance-swatches-pantone-color-1086726/

11:00 – 12:00 GMT
Characteristic Simple consent Informed consent
Type of decision Low risk High risk
Elements Explanation of
intervention, followed
by patient agreement
or refusal (expressed
or implied); other
elements, such as
discussion of risks,
benefits, and
alternatives are
present when
appropriate
Discussion of the
nature, purpose, risks
and benefits of
proposed intervention,
any alternatives, and
no treatment, followed
by explicit patient
agreement or refusal
Whitney, SN., McGuire, AL., & McCullough, LB 2004, ‘A typology of shared decision making, informed consent, and
simple consent’, Annals of Internal Medicine, vol.140, no. 1, pp. 54-59.

High
Low
RISK
Certain
(1 clear best choice)
CERTAINTY Uncertain
(˃ 2 alternatives)
Quadrant A: high risk (type of data/risk
of failure), high certainty
Consent type: Informed
Shared decision-making: Absent
Interaction: Intermediate, enough for an
adequately informed decision
Example: Financial aid, extended
programs
Quadrant C: low risk, high certainty
Consent type: Simple
Interaction: Minimal or none
Example: Lower diuretic dose for
patient with low serum potassium level
Quadrant D: low risk, low certainty
Shared decision-making: Present
Interaction: Intermediate
Example: Lifestyle changes vs
medication for hyperlipidaemia
Quadrant B: high risk, low certainty
Interaction: Extensive, including
discussion of patient values, preferences,
hopes and fears
Example: Mastectomy or lumpectomy
plus radiation for early breast cancer
Whitney, SN., McGuire, AL., & McCullough, LB 2004, ‘A typology of shared decision making, informed consent, and
simple consent’, Annals of Internal Medicine, vol.140, no. 1, pp. 54-59.

HighLow
RISK
Certain
(1 clear best choice) CERTAINTY
Uncertain
(˃ 2 alternatives)
Quadrant A: high risk (type of
data/risk of failure), high certainty
Interaction: Intermediate, enough for
an adequately informed decision
Example: Financial aid, extended
programs
Quadrant C: low risk, high certainty
Interaction: Minimal or none
Example: Use of aggregated data to
offer additional support to broad
cohorts, follow up support for missed
key milestones
Quadrant D: low risk, low certainty
Interaction: Intermediate
Example: Sending information
regarding tutorial classes, alternative
programs, reading material
Quadrant B: high risk, low certainty
Interaction: Extensive, including
discussion of student values,
preferences, hopes and fears
Example: Choice of degree program,
suggestion of a different institution,
ability to opt out of learning analytics
driven support
Source: Prinsloo, P., & Slade, S. (in press). Student consent in learning analytics: the devil in the details? In J.
Lester, C. Klein, H. Rangwala, and A. Johri (Eds), Learning analytics in higher education: Current innovations,
future potential, and practical applications. Routledge.

11:00 – 12:00 GMT
Elements of a more nuanced understanding and
praxis of student consent
• Simple versus informed consent
• The level of risk
• The certainty of our analysis, peer review, screening out of
biases and stereotypes
• Shared decision-making
• Different levels/intensity of interaction

11:00 – 12:00 GMT
A more nuanced understanding and praxis of student
consent will allow us to move beyond compliance
towards an ethical and data-informed praxis where
students are no longer passive recipients of services or
where they have no option to contest the outcome of
an analysis or contribute to a more informed position.
Image credit: https://www.quora.com/Whats-the-simplest-
way-to-create-an-SMS-opt-in-marketing-campaign
Image credit: http://chrisschweppe.com/?p=340

11:00 – 12:00 GMT
“Providing people with notice, access, and the ability to
control their data is key to facilitating some autonomy in
a world where decisions are increasingly made about
them with the use of personal data, automated
processes, and clandestine rationales, and where people
have minimal abilities to do anything about such
decisions”
(Solove, 2013, p. 1899; emphasis added)
(In)conclusions
Solove, D.J. 2013. Introduction: Privacy self-management and the consent dilemma. Harvard Law
Review, 1880 (2013); GWU Legal Studies Research Paper No. 2012-141; GWU Law School Public Law
Research Paper No. 2012-141. Available at SSRN: http://ssrn.com/abstract=2171018

11:00 – 12:00 GMT
Student consent in learning analytics:
Going into the labyrinth to face (tame?)
the Minotaur
Image credit: https://pixabay.com/en/maze-graphic-render-labyrinth-2264/
Alternative title

THANK YOU
Paul Prinsloo (Prof)
Research Professor in Open Distance Learning (ODL)
College of Economic and Management Sciences, Samuel Pauw
Building, Office 5-21, P.O. Box 392
Unisa, 0003, Republic of South Africa
T: +27 (0) 12 433 4719 (office)
prinsp@unisa.ac.za
Skype: paul.prinsloo59
Personal blog:
http://opendistanceteachingandlearning.wordpress.com
Twitter profile: @14prinsp
11:00 – 12:00 GMT

Student consent in learning analytics: Finding a way out of the labyrinth

More Related Content

What's hot (20)

Similar to Student consent in learning analytics: Finding a way out of the labyrinth (20)

More from University of South Africa (Unisa) (20)

Recently uploaded (20)

Student consent in learning analytics: Finding a way out of the labyrinth