Sign-Congruence, External Validity, and Replication
Political Analysis,
Год журнала:
2025,
Номер
unknown, С. 1 - 16
Опубликована: Янв. 17, 2025
Abstract
We
develop
a
formal
framework
for
accumulating
evidence
across
studies
and
apply
it
to
theoretical
foundations
replication.
Our
primary
contribution
is
characterize
the
relationship
between
replication
distinct
formulations
of
external
validity.
Whereas
conventional
wisdom
holds
that
facilitates
learning
about
validity,
we
show
this
not,
in
general,
case.
results
how
comparisons
magnitude
or
sign
empirical
findings
link
concepts
However,
without
careful
attention
research
design
constituent
studies,
can
mislead
efforts
assess
two
must
have
essentially
same
designs,
i.e.,
be
harmonized,
order
their
estimates
provide
information
any
kind
This
result
shows
even
minor
differences
study
its
introduce
discrepancy
typically
overlooked,
problem
becomes
more
pronounced
as
number
increases.
conclude
by
outlining
design-driven
approach
replication,
which
responds
issues
our
identifies
details
agenda
manage
them
productively.
Язык: Английский
Partisan Leaning
American Political Science Review,
Год журнала:
2025,
Номер
unknown, С. 1 - 20
Опубликована: Март 28, 2025
A
decisive
voter’s
exact
ideological
preferences
can
be
hard
to
predict,
even
for
seasoned
candidates.
We
develop
a
novel
theory
of
electoral
competition
where
candidates
are
evaluated
on
and
nonideological
dimensions.
The
key
feature
our
is
that
an
electorate’s
partisan
leaning
serves
as
signal
the
median
position
extreme
leanings
more
informative
about
voters
than
centrist
leanings.
show
this
leads
endogenous
sorting
districts
between
“extreme”
“centrist”
increase
in
importance
candidate
competence
increases
polarization—but
only
districts.
evaluate
using
data
from
mayoral
elections
Brazil’s
95
largest
municipalities
exploit
COVID-19
shock
salience
competence.
these
elections,
leading
increased
political
polarization,
which
concentrated
cities
with
Язык: Английский
External Validity and Evidence Accumulation
Опубликована: Ноя. 29, 2024
The
accumulation
of
empirical
evidence
that
has
been
collected
in
multiple
contexts,
places,
and
times
requires
a
more
comprehensive
understanding
research
than
is
typically
required
for
interpreting
the
findings
from
individual
studies.
We
advance
novel
conceptual
framework
where
causal
mechanisms
are
central
to
characterizing
social
phenomena
transcend
context,
place,
or
time.
distinguish
various
concepts
external
validity,
all
which
characterize
relationship
between
effects
produced
by
different
settings.
Approaches
require
careful
consideration
cross-study
features,
including
theoretical
considerations
link
constituent
studies
measurement
about
how
quantifed.
Our
main
contribution
developing
uniting
principles
constitute
qualitative
quantitative
assumptions
form
basis
then
apply
our
three
approaches
studying
general
phenomena:
meta-analysis,
replication,
extrapolation.
Язык: Английский