So what then should a code of ethics be?
What should it do?
Well of course we have to have an ethical approach to what we do, but we equally don't
want to put up prohibitive barriers that actually stop us doing the work that we want to do,
stop us doing sensible learning analytics.
So an ethics code needs to be a living, breathing document; something that is evolving all the
time.
Jisc put in place a code of practice at a very early stage of our journey in learning
analytics, but it is a living document, subject to amendments, something that we need to reinterpret
as our practice develops.
So what then are the major headings, the major subject areas that should be in a code of
practice?
So firstly responsibility.
Learning analytics collects sources of data from a wide range of departments within a
complex organisation like a university and a college.
People involved range from people who are at the rock face of collecting data, through
to people who analyse that data, record it and pass it on.
So a first key area for any code of ethics is to be very clear and transparent about
where that responsibility lies.
The code of ethics needs to identify who is responsible, define the extent of that responsibility,
and be quite clear and unambiguous about it, so that that responsibility can be mapped
on to a particular person's job responsibilities and professional obligations.
Transparency.
Well transparency is about openness and honesty.
We expect transparency in any public organisation or in any organisation that holds information
about us.
So transparency is about being very clear and open about what we're doing, so the person
that ultimately owns the data is very clear about where it's being used and where it's
going.
So a code of ethics should define what that means in terms of an institution.
How much are we making people aware of what we're doing, aware of our responsibilities,
aware of our obligations and their rights?
Consent.
Consent is getting more complicated.
So informed consent is essential for many legal processes in terms of data protection,
and is also generally accepted as an ethical thing to do.
So, in other words, we have to make sure that the students from whom we're collecting the
information have given a form of consent.
But consent is getting more and more legally complicated because there is a recognition
that consent is also about power.
So, for example, you can't really say that a person has given unconstrained consent if
a condition for them joining the university in the first place is they tick the box to
surrender their data.
So this is an area that's quite volatile and under consideration.
So we would expect a code of ethics translated into practice to certainly mention informed
consent.
But this is one of those areas that's alive and changing and we need to make sure that
we're on top of those changes as they develop.
Privacy.
Everybody has a right to privacy.
Now privacy here, when we're considering ethics and practice, needs to be differentiated from
de-identification.
So de-identification is really important when it comes to releasing personal information
to different levels within an organisation.
But de-identification is not in itself a substitute for privacy.
So when a student necessarily releases their personal data to their teacher, for example,
who has to have access to their individual sets of data, that teacher has a significant
number of privacy responsibilities in terms of both ethics and practice.
A lot of data is sensitive.
So, for example, a teacher might need to know about a person's disabilities.
But that right to know doesn't extend their right to share that information without consent
to others.
People need to be sensitive to that.
People need to be sensitive to the political environment on key indicators of identity,
for example, that can compromise a person in the world.
So if we're not ever mindful of privacy, we are likely to violate the trust that we have
with an individual, which is essential for learning analytics.
Validity.
Validity is about a cluster of issues.
It obviously includes accuracy; we want our information to be accurate.
Inaccurate information is simply wrong.
But it's more than just accuracy.
It's about recognising the danger of gaps in the data set, for example, which might
have implications.
It's about recognising that inappropriate correlations can be made that imply a cause-effect
relationship.
If those get into a system, then the learning analytics system lacks validity and we come
to false conclusions.
So validity extends the concept of accuracy into the process of analysis and the process
of reporting.
The reason why it's an ethical concern, that underlines a code of practice, is that invalid
data can have very severe consequences.
Very clearly an invalid data record can have very serious consequences for an individual
student.
If the mark is incorrect, the student can fail the degree.
So there's obviously a key interest in validity there.
But validity also extends to the public implications that we draw from information.
So at the opposite end of the spectrum, if we take something like the Teaching Excellence
Framework, for example, that depends on the validity of the data sets that originate in
individual institutions.
If those aren't valid then the public is going to draw quite wrong conclusions about the
implications of the relative value of different institutions, and potential students might
make inappropriate choices in what are some of the most important investments in their
lives.
So validity goes to the heart of the trust that we can expect in the system and the value
of that system both to individuals, organisations, but also to society at large.
Access.
Now remember that learning analytics depends on individual students surrendering their
rights over their personal information, and shifts the responsibility onto the organisation
to use that information appropriately.
So it follows from this that individual students have a right of access to how we're using
that data and the way that we're interpreting it.
And a code of ethics translated into a code of practice must allow for appropriate access.
It doesn't from this follow that a student will have access to everything that an organisation
knows about them or deduces about them.
For example, if the outcome of the analysis might be harmful for the student, the institution
might have a reason not to tell the student that that is their conclusion.
But the code of ethics must translate into practices that allow that to be done on an
everyday basis in a consistent, fair and reasonable manner.
Positive interventions.
Now remember that the key purpose of learning analytics is to improve the student experience
and to improve the learning environment; that word improvement runs through everything that
we do.
So it's ethical to expect that the application of learning analytics will lead to positive
outcomes.
Ethics, and the code of practice that follows from this, needs to ensure that we understand
what those are, and also that it defines what form those interventions will take, as we
move forward.
That will depend on the individual institution.
So the code of ethics will set that out in broad terms, but the individual institution
needs to define what form of interventions it's envisaging, how it's going to put them
in practice, and how always, those are going to lead to positive outcomes that are appropriate
to the mission of that institution in particular.
In turn, positive interventions depend on appropriate resourcing.
So you can't institute a programme of learning analytics, make promises to your students
and make promises to your staff, and then find that you have inadequate resources to
carry out those interventions.
Not only is that betrayal of trust, it's unethical, because promises have been made that can't
be met.
This is why it's essential that the decision to implement learning analytics is made at
the highest level of the institution, to ensure that appropriate resources are put behind
the interventions that will follow.
Don't promise students interventions based on their analytic results, that then as an
institution you can't meet.
Avoiding adverse impact.
This is a really important part of ethical practice that needs to be translated into
the way we behave, our codes of practice.
Clearly monitoring something always runs the risk of the intervention affecting the behaviour,
and whenever we're thinking about how we use learning analytics we've got to be very careful
that people don't respond by trying to game the system, or that we don't induce adverse
behaviour in people by the process of monitoring and reporting that we put in place.
This must be defined in terms of the particular level and the particular application of learning
analytics.
So, for example, if for our educators, our teachers, we set job performance goals that
require very, very specific outcomes, we mustn't create a situation where the use of learning
analytics distorts the value of the teaching simply to achieve the monitored outcomes.
At the organisational level where we have national indicators, for example, such as
the Teaching Excellence Framework, we need to avoid the tendency of institutions to drive
their whole policies towards achieving better scores on these sorts of narrow outcomes.
Now this is a very familiar problem across all aspects of education, whether we're talking
about the way we measure performance in primary schools, or the way that we measure performance
in universities.
And that difficulty, which we're well aware of, goes back into the heart of learning analytics,
and is why it has such an important part to play in a code of practice.
Stewardship is a really important part of any data driven organisation.
It also relates to data literacy.
It also relates to the whole concept of a data culture within in an institution.
Stewardship is our responsibility as the custodians of data.
It may not have originated with us.
It may not belong to us, but when we hold it we have a clear set of responsibilities
for it which extend to making sure that it is used appropriately, making sure it's accurate
and valid, making sure that it is not passed on inappropriately to third parties but equally
making sure that when necessary it is deleted or destroyed.
And those stewardship rules need to be very clearly set out within the code of ethics
and the code of practice.
There's naturally a tendency to think that data is owned or the responsibility of somebody
else, maybe somebody in the IT department or someone in the Registrar's department.
What data stewardship tells us is in our complicated organisations such as colleges and universities,
very many people are data stewards and we need to set out very clearly the ways in which
they should fulfil those responsibilities.
Overall, codes of ethical practice should be living, breathing documents that make sense
in the world.
They should be things that we think about that are sensible.
They shouldn't be prescriptive documents that stop us doing things.
They shouldn't be rules and they shouldn't be regulations.
We want an environment where we're free to innovate, where we're free to put in place
new systems, where we're free to do things that improve the quality of learning.
That's why it's so important that an ethical code is something that is actively discussed,
actively worked on, actively changed to changing circumstances.
At the end of the day it's a document that keeps us honest, keeps us on focus, keeps
us directed towards our primary purposes of doing good.
For more infomation >> Opel Corsa - Duration: 1:06. 
For more infomation >> Honda Jazz Elegance 1.4 i-VTEC CVT/Rijklaarprijs/ - Duration: 1:01. 
For more infomation >> Peugeot 208 Active 1.0 Vti 68 PK 5-DEURS | LAGE KILOMETERSTAND - Duration: 1:00. 
For more infomation >> Peugeot 208 5-deurs Active 1.6 BlueHDi 100 PK |LAGE KILOMETERSTAND| - Duration: 0:59. 
For more infomation >> Ultime notizie: Gossip Albano, Romina Power e Loredana Lecciso: la clamorosa decisione | K.N.B.T - Duration: 2:29. 
For more infomation >> Ford Ka+ 1.2 Trend Ultimate Cruisecontrol LM-velgen PDC - Duration: 0:59. 
For more infomation >> Opel Astra Wagon 1.7 CDTI ENJOY - Duration: 0:59. 
For more infomation >> Kirin J Callinan Bares All - The Feed - Duration: 8:53. 
For more infomation >> 💥Mon PIRE jeu 💥 Bonus 2000 abos ! - Duration: 4:06. 
For more infomation >> 박유천 성폭행 고소女, 결국 대법원 상고 취하 | 연예뉴스 24/7 - Duration: 3:08. 
For more infomation >> TOP 10 - První světová válka | šokující fakta - Duration: 4:08. 
No comments:
Post a Comment