A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Experimental Psychology General
This content is subject to copyright. Terms and conditions apply.
Journal
of
Experimental Psychology: General
1999, Vol.
128,
No.
3,
309-331
Copyright
1999
by the
American
Psychological
Association, Inc.
0096-3445/99/S3.00
Working
Memory,
Short-Term
Memory,
and
General Fluid Intelligence:
A
Latent-Variable
Approach
Randall
W.Engle
Georgia Institute
of
Technology
James
E.
Laughlin
University
of
South Carolina
Stephen
W.
Tuholski
Southern Illinois University
at
Edwardsville
Andrew
R. A.
Conway
University
of
Illinois
at
Chicago
A
study
was
conducted
in
which
133
participants performed
11
memory tasks (some thought
to
reflect working memory
and
some thought
to
reflect short-term memory),
2
tests
of
general
fluid
intelligence,
and the
Verbal
and
Quantitative
Scholastic
Aptitude
Tests.
Structural
equation modeling suggested that short-term
and
working memories reflect separate
but
highly related constructs
and
that many
of the
tasks used
in the
literature
as
working memory
tasks reflect
a
common construct. Working memory shows
a
strong connection
to fluid
intelligence,
but
short-term memory does not.
A
theory
of
working memory capacity
and
general
fluid
intelligence
is
proposed:
The
authors argue that working memory capacity
and
fluid
intelligence
reflect
the
ability
to
keep
a
representation active, particularly
in the
face
of
interference
and
distraction.
The
authors also discuss
the
relationship
of
this capability
to
controlled attention,
and the
functions
of the
prefrontal
cortex.
The
term short-term memory
is
often
used
to
refer
to a
concept
quite similar
to
working memory. (Anderson, 1990,
p.
150)
Short-term memory
is the
type
of
memory
we use
when
we
wish
to
retain information
for a
short time
to
think about
it.
The
short-term
store
has a
working memory component,
a
sort
of
mental workspace
or
sketchpad
in the
mind, that
is
used
to
manipulate information
in
consciousness.
(Seamon
&
Ken-
rick,
1994,
p.
220)
Working memory
is a
more complex construct than short-
term memory,
defined
as the set of
activated memory ele-
ments; there
is no
reason
to
doubt that working memory
is
based
on
that activated information along with central execu-
tive
processes.
(Cowan,
1995,
p.
100)
What
is one to
conclude
about
the
relationship between
working
memory (WM)
and
short-term
memory
(STM)
based
on the
preceding quotes?
The first
depicts
the two
constructs
as
"similar,"
the
second depicts
WM as a
subset
Randall
W.
Engle,
School
of
Psychology, Georgia Institute
of
Technology;
Stephen
W.
Tuholski, Department
of
Psychology,
Southern
Illinois
University
at
Edwardsville; James
E.
Laughlin,
Department
of
Psychology, University
of
South Carolina; Andrew
R. A.
Conway, Department
of
Psychology, University
of
Illinois
at
Chicago.
This
work
was
supported
by Air
Force
Office
of
Scientific
Research Grants
F49620-93-1-0336
and
F49620-97-1-0041
and
National Institute
of
Child Health
and
Human Development Grant
RO1-HD27490-01A1.
We
express
our
gratitude
to
Heidi Dickin-
son, Mike Kane, Natalie
Oransky,
Virginia Rosen,
and
Rebecca
Shisler
for
their help
in
collecting
the
data
and to
Timothy
Salthouse
and
Christopher
Hertzog
for
their
helpful
comments
on
the
manuscript.
Correspondence concerning this article should
be
addressed
to
Randall
W.
Engle,
School
of
Psychology, Georgia Institute
of
Technology,
Atlanta, Georgia
30332-0170.
Electronic mail
may be
sent
to
randall.engle@psych.gatech.edu.
of
STM,
and the
third depicts just
the
opposite:
that
STM is a
subset
of WM.
Ambiguity clearly exists
in the field of
cognitive
psychology regarding
the
relationship between
STM and WM.
Given
the
centrality
of
these concepts
to
grand
theories
of
cognition (e.g., Anderson, 1983; Cowan,
1995), remarkably little
work
has
been done
to
resolve this
confusion.
The
question addressed here
is "To
what
extent
are the
terms
STM and WM
different
terms
for the
same construct
and
to
what
extent
do
they
refer
to
different
but
more
or
less
related
constructs?"
Although
there
has
been considerable
experimental
work
on
each
of the
tasks that
putatively
reflect
the
two
concepts, little
work
has
been directed
at the
extent
to
which
those tasks share conceptual
and
construct
validity.
We
submit
that
these questions, although
very
important
to
mainstream
cognitive psychology, cannot
be
answered
by
experimental
studies alone,
but
require
a
combination
of
experimental
and
regression procedures.
Even
with
tasks
that
are
reported
to be WM
tasks (e.g., reading span,
operation
span,
computation
span,
n
back, etc.), there
is
little
solid
evidence
that
they,
in
fact,
reflect
a
common construct.
Even
if
they
do
reflect
a
single common construct,
we do not
know
whether
the
tasks measure processes
or
structures
that
are
different
from
those measured
by
tasks
that
putatively
reflect
STM
(e.g., simple digit
or
word span).
To
answer
the
questions
we set for
ourselves here,
we
analyzed
the
unique
and
shared variance across tasks
as
well
as the
underlying
factor
structure
of
that
variance
and the
extent
to
which
theoretical
conceptualizations
of the
underlying
constructs
are
supported
by
structural
models
of the
variance
in
those
tasks.
We
first
discuss
one
view
of the
nature
of STM and WM
and
justify
why we
think
both
constructs
are
necessary
for a
full
understanding
of
cognitive phenomena.
We
then
discuss
measurement concerns about these
two
constructs
and the
309
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.