now - Learner Analytics Summit

advertisement
Practical Ethical Models in an
Educational Age of Learning Analytics
James E. Willis, III, Ph.D.
Matthew D. Pistilli, Ph.D.
Educational Assessment Specialist
Research Scientist
Office of Institutional Research,
Assessment, and Effectiveness
Office of Institutional Research,
Assessment, and Effectiveness
July 2014
http://3.bp.blogspot.com/-Gv_qIJlLrWQ/UiZZP1FDU7I/AAAAAAAAAA8/FMhsNbO_2-E/s1600/ethics.jpg
Ethics in Learning Analytics
Ethics in learning analytics are just like
the rest of technology: right vs. wrong
It’s finally finding an audience.
•
• It’s about time.
Does binary approach really work?
What is the purpose of learning analytics?
(Slade and Prinsloo, 2011)
Academic Concerns
• Maximize number of
students reaching graduation.
• Improve completion rates
who may be disadvantaged.
Money
• Maximize profits.
Context: Ethics in LA Today
• “…systematization of correct and incorrect
behavior in virtual spaces according to all
stakeholders” (Pardo and Siemens, 2013, p. 2).
• Legal frameworks:
– “transparency, student control over the data, security,
and accountability and assessment” (Pardo and Siemens, 2013,
p. 11).
• Problem: Everything is based on utilitarianism
The Legal Connection
• Recent work (The Asilomar Convention for Learning Research in Higher Education)
affirmed six positions: “Respect for the rights and dignity of
learners, beneficence, justice, openness, the humanity of
learning, and continuous consideration.” Connected by the
1973 Code of Fair Information Practices and Belmont Report
of 1979.
• Utilitarianism and the law: both attempt to
provide a means to take individual
information and execute the greater good.
http://i.imgur.com/N7KOL.jpg
The Central Problem
• We cannot undue what has been done
• Technology = permanent possibility
Maybe the Problem is Modelling
• We tend to think in terms of modelling, both
scientifically and artistically.
• What happens if, when discussing ethics and LA, we
apply tensions instead?
• Modelling = preconditions for tenable outcomes
• Tensions = fluid, unfixed set of concepts to drive
questions and outcomes
Guiding Questions
• What leads us to ethical inquiry?
• What are the requisite problems that must be
considered, and why?
• What comes after ethical inquiry?
• Is there a post-ethic/s?
– How do we create the future by setting up how we
want to envisage it?
Moral Utopianism
• How the world ought to be, given perfect
circumstances.
• Technology: at worst, it does no harm; at best, it
provides better human, natural, and mechanized
environments.
• Learning Analytics: our technologies understand
what students need to learn and provide
calculated predictions to intervene meaningfully.
http://www.utopiamechanicus.com/wp-content/uploads/2011/12/umlogo.jpg
Moral Utopianism: Questions
• How do educators work out effective
interventions from past failures?
• How do we reform the values of failure into the
birthing pains of the future?
• How can learning analytics be a catalyst to steer
students productively in different directions?
Moral Ambiguity
• Value of an outcome cannot necessarily be
determined, and thus remains indefinitely
suspended (example: conflicting data and
undetermined directions).
• Technologically, this means actions may be taken
until there is legal precedent or public outcry
(example: Facebook, quantified self and insurance
companies).
• Learning Analytics: student ID cards, geo-tracking,
and grades correlation: what does this yield?
http://www.theplanningboardroom.net/wp-content/uploads/2010/04/ethics-pic1.jpg
http://4.bp.blogspot.com/-WdR8M2gNBU/T3hR9i47vDI/AAAAAAAAC74/soM9fEuPS9U/s1600/Image%2B%2Bmorality.jpg
http://ind.ccio.co/VB/F3/VA/futureWatchtypeSurveillancecamerasstrongreadtextmessagesfuturistic1.jpg
http://www.scientificamerican.com/sciam/cache/file/92885822-0614-48BF82279E8C3AD43A82.jpg
Moral Ambiguity: Questions
• What happens when the ‘nature’ of so-called big
data is repurposed or treated systematically?
How will educational institutions react?
• Our predictive analytics give us an entirely new
spectrum to change the future. No long in
conflict with individual agendas, learning
analytics can be pockets of shared data that
affect everyone. Are both proxies for ‘truth’?
Moral Nihilism
• Utter meaninglessness and lack of value;
nothing is intrinsically right or wrong,
including behavior.
• Technology: innovation may proceed without
any guidance or reflection because outcomes
have no value.
• Learning analytics: paternalism and care ethics
are rendered obsolete. Even retention for sake
of money is meaningless.
http://wolfweb.unr.edu/homepage/fenimore/ch202h/finalpapers/abacherli.jpg.jpg
http://www.redorbit.com/media/uploads/2012/07/tech-070812-005-617x416.jpg
Moral Nihilism: Questions
• Once administrators “know” something about a
student (via statistical regression), are institutions or
individuals compelled to act? What happens if there
is no action?
• What happens when something turns up in the data
(either as a single previously-unknown data point or
as a correlation of aggregate data) that is
unexpected? What infrastructure exists to handle it?
Building Ethics into Every Step
• Practical application is key
• We now stand on cusp of moral nihilism with
our technologies, including learning analytics
• Important: include probing questions,
assessments of possible outcomes, and active
disagreement about future developments
• Tensions: moral utopianism, ambiguity,
nihilism
Download