Human-centred learning analytics and AI in education: A systematic literature review
Computers and Education Artificial Intelligence,
Год журнала:
2024,
Номер
6, С. 100215 - 100215
Опубликована: Март 13, 2024
The
rapid
expansion
of
Learning
Analytics
(LA)
and
Artificial
Intelligence
in
Education
(AIED)
offers
new
scalable,
data-intensive
systems
but
raises
concerns
about
data
privacy
agency.
Excluding
stakeholders—like
students
teachers—from
the
design
process
can
potentially
lead
to
mistrust
inadequately
aligned
tools.
Despite
a
shift
towards
human-centred
recent
LA
AIED
research,
there
remain
gaps
our
understanding
importance
human
control,
safety,
reliability,
trustworthiness
implementation
these
systems.
We
conducted
systematic
literature
review
explore
gaps.
analysed
108
papers
provide
insights
i)
current
state
LA/AIED
research;
ii)
extent
which
educational
stakeholders
have
contributed
systems;
iii)
balance
between
control
computer
automation
such
iv)
reliability
been
considered
literature.
Results
indicate
some
consideration
system
design,
limited
end-user
involvement
actual
design.
Based
on
findings,
we
recommend:
1)
carefully
balancing
stakeholders'
designing
deploying
throughout
all
phases
2)
actively
involving
target
end-users,
especially
students,
delineate
automation,
3)
exploring
as
principles
future
Язык: Английский
Using learning analytics to enhance college students' shared epistemic agency in mobile instant messaging: A new way to support deep discussion
Journal of Computer Assisted Learning,
Год журнала:
2024,
Номер
40(3), С. 1166 - 1184
Опубликована: Янв. 17, 2024
Abstract
Background
Deep
discussions
play
an
important
role
in
students'
online
learning.
However,
researchers
have
largely
focused
on
engaging
students
deep
asynchronous
forums.
Few
studies
investigated
how
to
promote
discussion
via
mobile
instant
messaging
(MIM).
Objectives
In
this
study,
we
applied
learning
analytical
tools
(i.e.,
KBdeX
and
word
clouds)
enhance
shared
epistemic
agency
thereby
support
their
MIM.
Methods
Forty
Chinese
college
participated
study
reflected
MIM
engagement
by
participating
the
analytics
(LA)‐augmented
meta‐discourse
sessions.
The
used
multiple
data
analysis
methods,
including
content
analysis,
statistical
network
lag
sequential
analysis.
Results
We
found
that
LA
engaged
agency‐related
discourse,
such
as
creating
understanding,
knowledge
objects,
projective
regulative
processes.
particular,
clouds
more
complete
discourse
trajectory
which
started
from
awareness
of
unknowns,
then
progressed
setting
plans
sharing
information,
ultimately,
understanding.
Moreover,
our
indicated
moves
understanding
led
a
high
level
discussion.
Implications
This
contributes
research
extending
‘comparison
paradigm’,
focuses
comparing
(a)synchronous
forums
with
MIM,
‘design
mobilises
design
features
using
engage
promoting
agency.
Язык: Английский
Enhancing students’ dialogic reflection through classroom discourse visualisation
International Journal of Computer-Supported Collaborative Learning,
Год журнала:
2025,
Номер
unknown
Опубликована: Янв. 24, 2025
Abstract
Fostering
students’
dialogic
reflection
in
the
classroom
entails
promoting
diverse
perspectives
and
collaborative
aspects
of
reflective
thinking.
However,
few
studies
have
developed
visualisation
tools
to
create
a
environment
that
fosters
reflection,
especially
for
young
learners.
This
study
aimed
investigate
efficacy
visualising
discourse
enhance
reflection.
Two
information
communication
technology
(ICT)
teachers
their
students
(
n
=
53)
from
Hong
Kong
primary
school
participated
3-month
study.
After
engaging
visualisation-supported
subject
knowledge
ICT
improved,
accompanied
by
an
increase
proportion
words
spoken
during
lessons.
Interviews
with
two
eight
underscored
positive
impact
discourse,
which
enabled
embrace
alternative
problem-solving
approaches
actively
engage
both
group
work
whole-class
discussions.
illuminates
advantages
employing
as
tool.
Beyond
mere
revisiting,
can
view
experiences
multiple
reconstruct
knowledge.
Furthermore,
this
approach
fostered
accountability
contribution
interactions,
ultimately
improving
learning
outcomes.
These
findings
advocate
development
evidence-based,
specifically
designed
learners
optimise
experiences.
Язык: Английский
Improving participation equity in dialogic collaborative problem solving: A participatory visual learning analytical approach
Journal of Computer Assisted Learning,
Год журнала:
2024,
Номер
40(4), С. 1632 - 1657
Опубликована: Март 27, 2024
Abstract
Background
The
existing
research
on
dialogue‐based
learning
and
teaching
predominantly
highlights
its
capacity
to
yield
productive
educational
outcomes,
yet
it
often
overlooks
the
pivotal
factor
of
participation
equity,
which
is
fundamental
ensuring
efficacy
dialogic
learning.
Objectives
In
this
study,
equity
refers
a
condition
in
itself
opportunities
participate
are
fairly
distributed
among
participants
all
equally
listened
respected.
We
designed
technology‐enhanced
participatory
visual
analytical
approach
promote
equitable
collaborative
problem
solving
from
four
dimensions:
(i.e.,
equal
contributions
participants),
opportunity
promoting
engagement
with
others),
responsiveness
encouraging
attentiveness
others'
input),
respect
cultivating
respectful
communication
style).
Methods
intervention
class
fourth‐grade
students
(
n
=
59)
interacted
tool
reflect
their
learn
peer
talk
moves
address
issues
concerning
dimensions,
while
comparison
only
received
simple
feedback
respect.
Results
results
indicated
that
rates
group
members
were
insufficient
secure
equity.
was
effective
helping
realize
equity‐related
issues.
Intervention
more
regarding
than
students.
Conclusions
proposed
multidimensional
framework
has
potential
deepen
understanding
interactions.
Язык: Английский
Teaching talk for thinking: The efficacy of a peer talk teaching program for improving group thinking
Thinking Skills and Creativity,
Год журнала:
2023,
Номер
48, С. 101291 - 101291
Опубликована: Апрель 7, 2023
Язык: Английский
A Multi-Method Approach to Capture Quality of Collaborative Group Engagement
Computer-supported collaborative learning/The Computer-Supported Collaborative Learning Conference,
Год журнала:
2023,
Номер
unknown, С. 91 - 98
Опубликована: Окт. 3, 2023
Multi-method
approaches
are
an
emerging
trend
in
CSCL
research
as
they
allow
to
paint
a
more
comprehensive
picture
of
complex
group
learning
processes
than
using
single
method.In
this
contribution,
we
combined
measures
from
different
data
sources
capture
the
quality
collaborative
engagement
(QCGE)
CSCL-groups:
QCGE-self-assessments,
QCGE-ratings
verbal
communication,
and
video
recorded
nonverbal
behaviors.Using
methods
analysis,
visualized,
described,
analyzed
related
each
other.Here,
present
results
suggesting
that
interrelated:
For
instance,
behavior
(like
nodding
head)
is
high
communications.Results
preliminary
show
disparities,
too.Yet,
conclude
multi-method
approach
understanding
QCGE.Feasibility
suitability
discussed
conclusions
for
future
drawn.
Quality
Collaborative
Group
Engagement
how
measure
itSinha
et
al.
(
2015)
investigated
contexts
construct
consisting
four
dimensions
on
level.They
postulate
dimensions:
1)
Behavioral
relates
task-related
collaboration
defines
if
members
focus
common
task
be
solved
(i.e.,
on-task
behavior)
or
rather
get
distracted
off-task
includes
persistence
effort
investment.2)
Social
refers
cohesion
ways
communicating,
whether
groups
communicate
constructive
respectful
way
responsive
other.3)
Cognitive
addresses
development
mutual
about
proceed
well
joint
regulation
deep-level
strategies
learning.4)
Conceptualto-consequential
ability
interrelate
diverse
content
draw
meaningful
conclusions,
effectively
use
technological
conceptual
tools
develop
products.On
empirical
level,
Sinha
al.`s
(2015)
four-dimensional
conceptualization
their
respective
highlights
complexity
dynamics
those
aspects
varying
between
within
groups.The
authors
could
demonstrate
study
newly
developed
videobased
observational
assess
CSCL-groups
basically
quantifying
according
three
levels
(1=low,
2=moderate,
3=high)
with
low
clearly
differentiated
postulated
dimensions.Additionally,
demonstrated
comparative
case
studies
(by
qualitative
analyses)
certain
influenced
other
dimensions,
e.g.,
behavioral
social
fostered
cognitive
engagement,
which
then
facilitated
conceptual-to-consequential
engagement.These
highly
important
analyze
point
options
regulating
(e.g.,
socio-emotional
regulation,
cf.Hadwin
al.,
2017;Järvelä
&
Hadwin,
2013)
thus
find
support
it
practice
targeted
ways.However,
state,
further
methodological
developments
necessary
achieve
goal
investigate
nature
comprehensively
2023
Proceedings
©
ISLS
92
detail.In
next
section
provide
brief
overview
regarding
scope
CSCL-research
before
will
elaborate
our
own
multi-methods
have
QCGE.
Язык: Английский
Exploring the Trajectory of Learning and Using Productive Peer Talk Moves
Computer-supported collaborative learning/The Computer-Supported Collaborative Learning Conference,
Год журнала:
2023,
Номер
unknown, С. 193 - 196
Опубликована: Окт. 3, 2023
This
study
reports
how
pupils
learn
and
use
productive
peer
talk
moves
in
dialogic
collaborative
problem
solving
throughout
a
semester-long
teaching
program.The
results
revealed
three-stage
trajectory:
intrusive
application,
structured
autonomous
application.It
takes
time
practice
for
students
to
appropriate
benefit
from
their
usage.Using
was
not
facilitative
the
problem-solving
process
at
beginning,
even
disruptive.Students
gradually
were
able
internalize
then
could
integrate
with
own
semantic
expressive
intentions.The
learning
of
also
aroused
cascade
long-term
effects,
including
more
respectful
language
styles,
shift
personal
pronouns
"I"
"we",
enrichment
body
gestures.
Язык: Английский
Human-Centred Learning Analytics and AI in Education: a Systematic Literature Review
arXiv (Cornell University),
Год журнала:
2023,
Номер
unknown
Опубликована: Янв. 1, 2023
The
rapid
expansion
of
Learning
Analytics
(LA)
and
Artificial
Intelligence
in
Education
(AIED)
offers
new
scalable,
data-intensive
systems
but
also
raises
concerns
about
data
privacy
agency.
Excluding
stakeholders
--
like
students
teachers
from
the
design
process
can
potentially
lead
to
mistrust
inadequately
aligned
tools.
Despite
a
shift
towards
human-centred
recent
LA
AIED
research,
there
remain
gaps
our
understanding
importance
human
control,
safety,
reliability,
trustworthiness
implementation
these
systems.
We
conducted
systematic
literature
review
explore
gaps.
analysed
108
papers
provide
insights
i)
current
state
LA/AIED
research;
ii)
extent
which
educational
have
contributed
systems;
iii)
balance
between
control
computer
automation
such
iv)
reliability
been
considered
literature.
Results
indicate
some
consideration
system
design,
limited
end-user
involvement
actual
design.
Based
on
findings,
we
recommend:
1)
carefully
balancing
stakeholders'
designing
deploying
throughout
all
phases,
2)
actively
involving
target
end-users,
especially
students,
delineate
automation,
3)
exploring
as
principles
future
Язык: Английский
Visual Learning Analytics for Educational Interventions in Primary and Secondary Schools
Journal of Learning Analytics,
Год журнала:
2024,
Номер
11(2), С. 91 - 111
Опубликована: Июнь 30, 2024
Visual
Learning
Analytics
(VLA)
uses
analytics
to
monitor
and
assess
educational
data
by
combining
visual
automated
analysis
provide
explanations.
Such
tools
could
aid
teachers
in
primary
secondary
schools
making
pedagogical
decisions,
however,
the
evidence
of
their
effectiveness
benefits
is
still
limited.
With
this
scoping
review,
we
a
comprehensive
overview
related
research
on
proposed
VLA
methods,
as
well
identifying
any
gaps
literature
that
assist
describing
new
helpful
directions
field.
This
review
searched
all
relevant
articles
five
accessible
databases
—
Scopus,
Web
Science,
ERIC,
ACM,
IEEE
Xplore
using
40
keywords.
These
studies
were
mapped,
categorized,
summarized
based
objectives,
collected
data,
intervention
approaches
employed,
results
obtained.
The
determined
what
affordances
allowed,
kind
visualizations
used
inform
students,
and,
more
importantly,
positive
interventions.
We
conclude
there
are
moderate-to-clear
learning
improvements
within
limit
studies’
interventions
support
use
tools.
More
systematic
needed
determine
whether
gains
translated
into
long-term
improvements.
Язык: Английский
Development of Visual Learning Analytic Tools to Explore Performance and Engagement of Students in Primary, Secondary, and Higher Education
Опубликована: Авг. 20, 2024
Schools
and
educational
institutions
collect
large
amounts
of
data
about
students
their
learning,
including
text,
grades,
quizzes,
timestamps,
other
activities.
However,
in
primary
secondary
education,
this
is
often
dispersed
across
different
digital
platforms,
lacking
standardized
methods
for
collection,
processing,
analysis,
presentation.
These
issues
hinder
teachers
from
making
informed
decisions
or
strategic
effective
use
data.
This
presents
a
significant
obstacle
to
progress
education
the
development
Educational
Technology
(EdTech)
products.
Visual
Learning
Analytics
(VLA)
tools,
also
known
as
Dashboards
(LADs),
are
designed
visualize
student
support
pedagogical
decision-making.
Despite
potential,
effectiveness
these
tools
remains
limited.
Addressing
challenges
requires
both
technical
solutions
thoughtful
design
considerations,
explored
Papers
1
through
5
thesis.
Paper
examines
aspects
VLA
by
evaluating
higher
various
visualization
Machine
(ML)
techniques.
2
provides
broader
insights
into
landscape
systematic
review,
mapping
key
concepts
research
gaps
emphasizing
potential
enhance
learning
outcomes.
Meanwhile,
3
delves
solution
(data
pipeline
standard)
considering
secure
Swedish
warehouse,
SUNET.
includes
standard
integrating
SUNET,
along
with
customized
scripts
reformat,
merge,
hash
multiple
datasets.
4
focus
on
aspects,
discussing
proposed
Human-Centered
Design
(HCD)
approach
involving
co-designing
simple
tool.
introduces
scenario-based
framework
Multiple
(MLADs)
development,
stressing
user
engagement
tailored
LADs
that
facilitate
decision-making
education.
The
dissertation
offers
comprehensive
advancing
user-centric
principles.
By
addressing
integration
users
tool
efforts
aim
empower
leveraging
improved
teaching
experiences.
Язык: Английский