Cognition,
Journal Year:
2025,
Volume and Issue:
258, P. 106088 - 106088
Published: Feb. 21, 2025
The
sensory
input
that
we
encounter
while
navigating
through
each
day
is
highly
structured,
containing
patterns
repeat
over
time.
Statistical
learning
the
process
of
becoming
attuned
to
these
and
can
facilitate
online
processing.
These
facilitation
effects
are
often
ascribed
prediction,
in
which
information
about
an
upcoming
event
represented
before
it
occurs.
However,
previously
observed
could
also
be
due
retrospective
Here,
using
a
speech-based
segmentation
paradigm,
tested
whether
statistical
leads
prediction
syllables.
Specifically,
probed
for
behavioural
hallmark
genuine
given
benefits
processing
when
confirmed,
but
incurs
costs
if
disconfirmed.
In
line
with
idea
key
outcome
learning,
found
trade-off
greater
benefit
predictable
syllables
was
associated
cost
occurred
"mismatch"
context,
outside
their
expected
positions.
This
making
predictions
evident
at
both
participant
item
(i.e.,
individual
syllable)
level.
Further,
did
not
emerge
indiscriminately
all
stream,
deployed
selectively
according
trial-by-trial
demands
task.
Explicit
knowledge
word
required
occur,
suggesting
operates
largely
implicitly.
Overall,
results
provide
novel
evidence
arises
as
natural
consequence
learning.
Abstract
In
an
auditory
statistical
learning
paradigm,
listeners
learn
to
partition
a
continuous
stream
of
syllables
by
discovering
the
repeating
syllable
patterns
that
constitute
speech
stream.
Here,
we
ask
whether
benefits
from
spaced
exposure
compared
with
massed
exposure.
longitudinal
online
study
on
Prolific,
exposed
100
participants
regularities
in
way
(i.e.,
blocks
spread
out
over
3
days)
and
another
all
lumped
together
single
day).
phase,
listened
streams
composed
pairs
while
responding
target
syllable.
The
groups
exhibited
equal
during
exposure,
as
indicated
comparable
response-time
advantage
for
predictable
syllables.
However,
terms
resulting
long-term
knowledge,
observed
benefit
Following
2-week
delay
period,
tested
participants’
knowledge
forced-choice
test.
While
both
performed
above
chance,
group
had
higher
accuracy.
Our
findings
speak
importance
timing
structured
input
also
outside
laboratory
(e.g.,
language
development),
imply
current
investigations
likely
underestimate
human
abilities.
Cerebral Cortex,
Journal Year:
2025,
Volume and Issue:
35(2)
Published: Feb. 1, 2025
Extracting
spatial
or
temporal
patterns
across
experiences
is
essential
for
skill
acquisition
and
predictive
processes.
The
prefrontal
cortex
plays
a
central
role
in
regulating
competitive
cognitive
systems,
with
particular
influence
on
executive
functions,
often
opposing
statistical
learning.
This
regulatory
function
may
account
observed
improvements
the
consolidation
of
regularities
following
inhibition
dorsolateral
via
repetitive
transcranial
magnetic
stimulation.
However,
whether
access
to
previously
acquired
knowledge
can
similarly
benefit
from
remains
unclear.
preregistered
study
investigated
cortex's
retrieving
pre-existing
regularities.
Healthy
human
participants
engaged
an
implicit
probabilistic
sequence
learning
task
followed
by
24-h
period.
Before
retesting,
they
received
either
1
Hz
stimulation
sham
over
left,
right,
bilateral
10
min.
We
that
retrieval
was
enhanced
Bilateral
group
compared
Sham
group.
Our
findings
suggest
facilitate
knowledge,
particularly
when
interhemispheric
compensatory
mechanisms
are
limited.
These
insights
advance
our
understanding
dynamic
neural
background
inform
strategies
enhancement.
Cognition,
Journal Year:
2025,
Volume and Issue:
258, P. 106088 - 106088
Published: Feb. 21, 2025
The
sensory
input
that
we
encounter
while
navigating
through
each
day
is
highly
structured,
containing
patterns
repeat
over
time.
Statistical
learning
the
process
of
becoming
attuned
to
these
and
can
facilitate
online
processing.
These
facilitation
effects
are
often
ascribed
prediction,
in
which
information
about
an
upcoming
event
represented
before
it
occurs.
However,
previously
observed
could
also
be
due
retrospective
Here,
using
a
speech-based
segmentation
paradigm,
tested
whether
statistical
leads
prediction
syllables.
Specifically,
probed
for
behavioural
hallmark
genuine
given
benefits
processing
when
confirmed,
but
incurs
costs
if
disconfirmed.
In
line
with
idea
key
outcome
learning,
found
trade-off
greater
benefit
predictable
syllables
was
associated
cost
occurred
"mismatch"
context,
outside
their
expected
positions.
This
making
predictions
evident
at
both
participant
item
(i.e.,
individual
syllable)
level.
Further,
did
not
emerge
indiscriminately
all
stream,
deployed
selectively
according
trial-by-trial
demands
task.
Explicit
knowledge
word
required
occur,
suggesting
operates
largely
implicitly.
Overall,
results
provide
novel
evidence
arises
as
natural
consequence
learning.
