Article

Feature binding and attention in working memory: A resolution of previous contradictory findings

a Institute of Psychological Sciences , University of Leeds , Leeds , UK.
Quarterly journal of experimental psychology (2006) (Impact Factor: 2.13). 06/2012; 65(12):2369-2383. DOI: 10.1080/17470218.2012.687384
Source: PubMed

ABSTRACT

We aimed to resolve an apparent contradiction between previous experiments from different laboratories, using dual-task methodology to compare effects of a concurrent executive load on immediate recognition memory for colours or shapes of items or their colour–shape combinations. Results of two experiments confirmed previous evidence that an irrelevant attentional load interferes equally with memory for features and memory for feature bindings. Detailed analyses suggested that previous contradictory evidence arose from limitations in the way recognition memory was measured. The present findings are inconsistent with an earlier suggestion that feature binding takes place within a multimodal episodic buffer Baddeley, (20006.

Baddeley , A. D. 2000. The episodic buffer: A new component of working memory?. Trends in Cognitive Sciences, 4(11): 417–423. (doi:10.1016/S1364-6613(00)01538-2) [CrossRef], [PubMed], [Web of Science ®], [CSA]View all references) and support a subsequent account in which binding takes place automatically prior to information entering the episodic buffer Baddeley, Allen, & Hitch, (20117.

Baddeley , A. D. ,
Allen , R. J. and
Hitch , G. J. 2011. Binding in visual working memory: The role of the episodic buffer. Neuropsychologia, 49: 1393–1400. (doi:10.1016/j.neuropsychologia.2010.12.042) [CrossRef], [PubMed], [Web of Science ®]View all references). Methodologically, the results suggest that different measures of recognition memory performance (A′, d′, corrected recognition) give a converging picture of main effects, but are less consistent in detecting interactions. We suggest that this limitation on the reliability of measuring recognition should be taken into account in future research so as to avoid problems of replication that turn out to be more apparent than real.

Full-text

Available from: Judit Castellà
This article was downloaded by: [University of Leeds]
On: 23 June 2012, At: 04:39
Publisher: Psychology Press
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office:
Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
The Quarterly Journal of Experimental
Psychology
Publication details, including instructions for authors and subscription
information:
http://www.tandfonline.com/loi/pqje20
Feature binding and attention in working
memory: A resolution of previous
contradictory findings
Richard J. Allen
a
, Graham J. Hitch
b
, Judit Mate
c
& Alan D. Baddeley
b
a
Institute of Psychological Sciences, University of Leeds, Leeds, UK
b
Department of Psychology, University of York, York, UK
c
Departament de Psicologia Bàsica, Evolutiva i de l'Educacio, Universitat
Autònoma de Barcelona, Barcelona, Spain
Available online: 30 Apr 2012
To cite this article: Richard J. Allen, Graham J. Hitch, Judit Mate & Alan D. Baddeley (2012): Feature
binding and attention in working memory: A resolution of previous contradictory findings, The Quarterly
Journal of Experimental Psychology, DOI:10.1080/17470218.2012.687384
To link to this article: http://dx.doi.org/10.1080/17470218.2012.687384
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial
or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the
contents will be complete or accurate or up to date. The accuracy of any instructions, formulae,
and drug doses should be independently verified with primary sources. The publisher shall not
be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or
howsoever caused arising directly or indirectly in connection with or arising out of the use of this
material.
Page 1
Feature binding and attention in working memory:
A resolution of previous contradictory ndings
Richard J. Allen
1
, Graham J. Hitch
2
, Judit Mate
3
, and Alan D. Baddeley
2
1
Institute of Psychological Sciences, University of Leeds, Leeds, UK
2
Department of Psychology, University of York, York, UK
3
Departament de Psicologia Bàsica, Evolutiva i de lEducacio, Universitat Autònoma de Barcelona, Barcelona,
Spain
We aimed to resolve an apparent contradiction between previous experiments from different laboratories,
using dual-task methodology to compare effects of a concurrent executive load on immediate recognition
memory for colours or shapes of items or their colourshape combinations. Results of two experiments
conrmed previous evidence that an irrelevant attentional load interferes equally with memory for fea-
tures and memory for feature bindings. Detailed analyses suggested that previous contradictory evidence
arose from limitations in the way recognition memory was measured. The present ndings are incon-
sistent with an earlier suggestion that feature binding takes place within a multimodal episodic buffer
(Baddeley, 2000) and support a subsequent account in which binding takes place automatically prior
to information entering the episodic buffer (Baddeley, Allen, & Hitch, 2011). Methodologically, the
results suggest that different measures of recognition memory performance (A
,d
, corrected recognition)
give a converging picture of main effects, but are less consistent in detecting interactions. We suggest that
this limitation on the reliability of measuring recognition should be taken into account in future research
so as to avoid problems of replication that turn out to be more apparent than real.
Keywords: Working memory; Binding; Attention; Recognition.
When we encounter objects comprising multiple fea-
tures such as shape, colour, orientation, or location,
what kind of representations are created in working
memory, and what processing and storage resources
contribute to this? It is clear that the different features
of an object can be temporarily bound or integrated,
with participants able to respond successfully to a
range of measures that indicate memory not only
for individual features but for feature conjunctions
(e.g., Allen, Baddeley, & Hitch, 2006; Brown &
Brockmole, 2010; Delvenne, Cleeremans, &
Laloyaux, 2010; Gajewski & Brockmole, 2006;
Johnson, Hollingworth, & Luck, 2008; Logie,
Brockmole, & Vandenbroucke, 2009; Wheeler &
Treisman, 2002). However, debate has emerged in
the literature regarding the ways in which features
and their associations are stored and subsequently
displaced from memory, and whether attentional
Correspondence should be addressed to Richard J. Allen, Institute of Psychological Sciences, University of Leeds, LS2 9JT, UK.
E-mail: r.allen@leeds.ac.uk
We would like to thank Louise Brown for useful discussion, and Geoffrey Woodman and James Brockmole for helpful comments
on an earlier version of this manuscript.
# 2012 The Experimental Psychology Society 1
http://www.psypress.com/qjep http://dx.doi.org/10.1080/17470218.2012.687384
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY
2012, iFirst,115
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 2
resources (both spatial and executive) are key to these
processes.
One important issue concerns whether visual
working memory is structured primarily in terms
of objects or features. For example, Luck and
Vogel (1997; Vogel, Woodman, & Luck, 2001)
argued for an object-based limit to visual working
memory, with features (e.g., shape, colour) stored
together as bound object representations (i.e.,
coloured shapes; see also Delvenne & Bruyer,
2004; Gajewski & Brockmole, 2006; Xu, 2002),
an argument that ts with research suggesting
object-based spatial attention (e.g., Duncan,
1984). Conversely, Wheeler and Treisman (2002)
suggested that different feature dimensions such
as shape, colour, orientation, and location might
be held in separable stores, with the main limit
on storage being the number of features from the
same dimension. A third possibility is that both
feature- and objec t-level resources are available to
support visual working memory (Baddeley, Allen,
& Hitch, 2011; Fougnie, Asplund, & Marois,
2010; Olson & Jiang, 2002; Parra, Della Sala,
Logie, & Abrahams, 2009; Ueno, Allen,
Baddeley, Hitch, & Saito, 2011).
An important and related question conce rns
the role of attentional control in processing features
and their conjunctions. In proposing the multi-
modal episodic buffer as a further component of
working memory, additional to the original
Baddeley and Hitch (1974) tripartite structure of
specialized phonological and visuospatia l subsys-
tems and a central executive, Baddeley (2000)
suggested that the initial creation of feature bind-
ings takes place within the episodic buffer and is
crucially dependent upon executive-based atten-
tional support. In line with this, vanRullen (2009)
has argued that, in contrast to processing of fam-
iliar, hard-wired conjunctions, the creation of
on-demand arbitrarily paired feature bindings in
object recognition particularly requires attention.
Similarly, a central element of the Wheeler and
Treisman (2002) feature-based approach is that
the bindings or connections between features
require focused attentional control to stay intact,
and that these bindings dissipate when attention
is disrupted or withdrawn. Allen et al. (2006)
explored these possibilities by examining whether
memory for shapes, colours, and shapecolour
bindings was differentially disrupted by the
addition of an attention-demanding verbal task
(e.g., backward counting) during encoding and
retention of to-be-remembered stimuli, with the
logic that any condition that is particularly reliant
on executive attention should show greater dis-
ruption. Substantial attentional load effects were
observed in all three conditions, indicating a clear
role for executive resources in visual working
memory generally (see also DellAcqua &
Jolicoeur, 2000; Morey & Cowan, 2004, 2005;
Stevanovski & Jolicoeur, 2007). However, these
effects were equivalent in magnitude for single-
feature and for feature-binding conditions, which
was taken to indicate that the binding of simple
visual features in working memory emerges relatively
automatically, possibly through perceptual and
visuospatial working memory processes, with the
results of this binding process then feeding into
the episodic buffer and conscious awareness
(Baddeley et al., 2011). Research has also indicated
that the subsequent maintenance of feature conjunc-
tion information following encoding is also not cri-
tically dependent on focused attention (Delvenne
et al., 2010; Gajewski & Brockmole, 2006;
Johnson et al., 2008; Yeh, Yang, & Chiu, 2005;
but see Fougnie & Marois, 2009), though bound
object representations may be by their nature
fragile and susceptible to interference (Allen et al.,
2006; Gorgoraptis, Catalao, Bays, & Husain,
2011; Logie et al., 2009; Ueno, Allen, et al., 2011;
Ueno, Mate, Allen, Hitch, & Baddeley, 2011).
However, it has recently been suggested that
visual feature binding may not be as automatic as
previously argued. For example, Elsley and
Parmentier (2009) observed that the binding of
verbal information (letters) to spatial locations
was disrupted when a demanding tone memory
task was performed concurrently, leading them to
reject the argument that binding is automatic.
However, verbalspatial conjunctions are cross-
domain in nature and potentially involve different
processes from the binding of visual features
within un itized objects. It is also worth noting
that concurrent task responses were collected after
2 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0)
ALLEN ET AL.
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 3
memory judgements, and therefore that the tone
maintenance task is likely to have disrupted the
response phase (cf. Allen et al., 2006; Allen,
Hitch, & Baddeley, 2009; Karlsen, Allen,
Baddeley, & Hitch, 2010).
In a study more direct ly relevant to the form of
binding presently under consideration, Brown and
Brockmole (2010) examined memory for shapes,
colours, and shapecolour conjunctions in young
and older adults. They found that, for both age
groups, concurrent backward counting perform-
ance had a signicantly greater disruptive effect
on their binding condition than it did on single fea-
tures. This would seem to challenge the notion that
bindings are created automatically and that execu-
tive control is not crucial in this process (e.g.,
Allen et al., 2006; Baddeley et al., 2011). It is there-
fore important to investigate this conicting
pattern of ndings, paying particular attention to
methodological differences between Allen et al.
(2006) and Brown and Brockmole (2010), several
of which were indeed noted in the latter study.
Firstly, and in common with Elsley and
Parmentier (2009), the backward counting task in
Brown and Brockmole (2010) was performed
throughout all stages (that is, presentation, delay,
and test) of the visual memory task, while Allen
et al. (2006) only required counting to continue
during presentation and delay. It is possible that
the recognition judgement in the binding con-
dition, requiring a decision on whether a test
probe pairing of features was encountered together
originally, is more difcult and/or attention
demanding than the feature test judgements. It
could be this component of task performance
(rather than binding encoding or maintenance
per se) that is more disrupted by a concurrent
executive load (Baddeley et al., 2011). Indeed,
recent research has indicated that underlying pro-
cesses may vary across different forms of recog-
nition and recall test, depending on factors such
as sampletest similarity and the type of decision
required at test (e.g., Alvarez & Thompson,
2009; Awh, Barton, & Vogel, 2007; Hyun,
Woodman, Vogel, Hollingworth, & Luck, 2009).
To explore whether additional complexity at the
test phase is responsible for effects on binding,
participants in the present studies were asked to con-
tinue performing the easy or more difcult concur-
rent verbal tasks during the presentation, delay, and
recognition judgement phases of each memory trial.
Secondly, Brown and Brockmole (2010) pre-
sented their three-item stimulus arrays for 900 ms,
providing considerably longer encoding time than
the 250 ms that was available for the four items in
Allen et al. (2006). It is possible that feature
binding is automatically achieved on the basis of per-
ceptual processing when items are encountered
briey, enabling the construction of representations
that are adequate for subsequent task performance.
In contrast, longer exposures may enable the appli-
cation of consciously controlled elaborative
binding mechanisms (Mitchell, Johnson, Raye,
Mather, & DEsposito, 2000). Thus, it is possible
that the experiments in Allen et al. (2006) tapped
automatized perceptual binding, rather than a
slower and more demanding form of episodic
binding as described by Baddeley (2000). In order
to check this possibility, the present studies used a
presentation duration of 1,000 ms throughou t.
Typically, recognition judgements are more
accurate for colour than for shape or colourshape
binding (e.g., Allen et al., 2006; Brown &
Brockmole, 2010; Song & Jiang, 2006; Wheeler
& Treisman, 2002). It is notable that, using
three-item displays, Brown and Brockmole (2010)
encountered very high performance levels in their
colour condition, which were near or at ceiling.
This has the potential to articially limit the size
of any concurrent task effect in this condition,
meaning that valid comparisons of relative interfer-
ence effects are only possible between shape and
binding (and not shape vs. colour, or colour vs.
binding). However, it is important to compare
the binding condition against both shape and
colour, if we are to condently identify binding-
specic phenomena. Therefore, in an attempt to
avoid ceiling effects when testing memory for
colour, and also to match the array sizes used by
Brown and Brockmole (2010) and Allen et al.
(2006), we varied set size in the present studies,
directly comparing three- and four-item displays.
A further difference apparent between the
studies under consideration concerns the form in
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0) 3
BINDING AND ATTENTION
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 4
which the data were reported and analysed. Allen
et al. (2006) focused on corrected recognition (hits
minus false alarms), noting that d
produced equiv-
alent patterns of data, while signicant variations in
response bias between experimental conditions ren-
dered A
inaccurate as a measure of sensitivity
(McNicol, 1972; Pastore, Crawley, Berens, &
Skelly, 2003; Snodgrass & Corwin, 1988). In con-
trast, but in line with a number of other studies
using recognition-based me asures of visual
memory (e.g., Allen et al., 2009; Delvenne &
Dent, 2008; Karlsen et al., 2010; Morey, 2011;
Xu, 2006), Brown and Brockmole (2010) reported
their data in terms of A
(Pollack & Norman,
1964). As corrected recognition and different
methods of signal detection can sometimes
provide different patterns of output (e.g., Pastore
et al., 2003), it is therefore of value to report and
analyse our ndings in the present study using
each of these data formats, both to provide a com-
prehensive overview of the ndings, thus helping
address the core issues under debate, and also to
provide potentially useful insights into scoring
methodology more generally.
We report two experiments examining memory
for shape, colour, and shapecolour binding, with
participants performing either a simple (repetition
of two-digit numbers) or an attention-demanding
verbal task (backward counting in decrements of
three using two-digit numbers) during the presen -
tation of stimuli, their maintenance over a short
delay, and the subsequent recognition probe test
phase. If the binding condition is particularly
reliant on central executive resources during any of
these phases, then it should show more disruption
as a result of the backward counting task, relative
to both the shape and colour conditions in isolation.
EXPERIMENT 1
Method
Participants
There were 24 participants (6 males and 18
females, aged between 19 and 23 years, M =
20.54, SD = 1.02) in this experiment. All were
undergraduate students at the University of York,
had English as a rst language, and showed
normal or corrected-to-normal vision.
Materials
All stimuli and probe items measured approxi-
mately 1.6 cm
2
and were presented on a grey back-
ground. The stimulus pool consisted of sets of eight
colours (blue, brown, green, purple, red, turquoise ,
white, yellow) and eight shapes (arch, chevron,
circle, cross, diamond, ag, star, triangle). Shape
outlines and RGB colour values are provided in
Figure 1. Combinations of these shapes and
colours were used as stimuli and test probes in
the binding condition. In contrast, in the colour
condition, stimuli were presented and tested using
a noncanonically shaped blob, while the shape
condition always involved black shape outlines
lled grey to match the background. Thus, as in
Allen et al. (2006), the irrelevant feature dimension
was always held constant during each of the feature
conditions. All stimuli were sampled randomly
without replacement within each trial.
Design and procedure
The experiment implemented a 3 × 2 × 2 repeated
measures design, manipulating stimulus condition
(colour; shape; binding), concurrent task (articula-
tory suppression, AS; backward counting, BC),
and set size (using displays of 3 and 4 items).
Conditions were divided into six blocks, based on
stimulus condition and concurrent task. Within
each of these blocks, there were 6 practice trials
(all at set size 3) and 48 test trials. The rst 24
test trials always featured three-item displays,
while the second 24 used four-item displays.
Suppression and counting blocks were yoked so
that all of the stimulus conditions for each concur-
rent task were performed together. Condition order
was counterbalanced across participants.
Trial procedure is illustrated in Figure 1. Each
trial started with the presentation of a randomly
generated two-digit number (between 20 and 99)
in the centre of the screen for 2 s. Participants
were instructed to either repeatedly articulate this
number out loud (AS) or count backwards in decre-
ments of 3 from it (BC), and to continue doing so
4 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0)
ALLEN ET AL.
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 5
until after they had made their key press recog-
nition judgement in response to the primary
memory task. Backward counting responses were
recorded by the experimenter.
The two-digit counting number was then
replaced with a xation cross in the upper centre
of the screen for 500 ms, followed by a 250-ms
blank screen delay and then the to-be-rememb ered
array. These arrays consisted of stimuli (colour
blobs in the colour condition, shape outlines in
the shape condition, and coloured shapes in the
binding condition) presented in a horizontal row,
with approximately 1.5 cm between each stimulus.
Both the three-item and the four-item displays
were centred at the location previously occupied
by the xation cross. Items remained on screen
for 1 s and were followed by a 1-s blank screen
retention interval.
The test probe was then displayed at lower
screen centre, with participants required to judge
whether this colour, shape, or colourshape combi-
nation had been present in the original array. For
the single-feature conditions, 50% of trials involved
a shape (or colour) probe that was present on that
trial, while 50% were lure trials involving a feature
from the wider experimental set that had not
been present. In the binding condition, in contrast,
the features of the test probe had always been
present in the original array, but were either re-
presented in their original combination, or drawn
from two different objects. Thus, accurate perform-
ance required memory for shapecolour conjunc-
tions in the latter condition but only either
shape or colour in the single-feature conditions.
Target and lure trials were randomly intermixed
within each block. The probe remained visible
until participants made their response via a key
press (using z key for yes responses or / for
no responses in all conditions), with accuracy
emphasized over speed.
Figure 1. (a) Trial sequence, using a recombination trial from the binding condition as an example. (b) Sample presentation displays from
colour and shape conditions in Experiment 1. In Experiment 2, all conditions used displays of coloured shapes (as in binding condition).
Note, grey shades denote different colours, background displays were grey, illustrations not to scale. (c) Set of shapes used in each experiment.
Colour values (RGB): black (0, 0, 0); blue (0, 0, 254); green (0, 255, 1); purple (201, 0, 200); red (254, 0, 0); turquoise (1, 255, 255);
white (255, 255, 255); yellow (255, 255, 1); grey background (171, 171, 171).
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0) 5
BINDING AND ATTENTION
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 6
Results
There were no signicant main effects or
interactions for either number of BC responses
(M = 3.63, SE = 15), p .28, or BC error rates
(M = .20, SE = .04), p .28.
Recognition judgement accuracy is reported as
corrected recognition (hits false alarms), d
, and
A
(see Stanislaw & Todorov, 1999). The outcomes
of these response measures are displayed in Table 1,
with corrected recognition also displayed in Figure
2. Each measure is analysed separately in a 3 × 2 ×
2 repeated measures ANOVA.
Starting with corrected recognition, we observed
signicant effects of stimulus condition, F(2, 46) =
60.08, MSE = .05, p , .001, η
2
p
= .72, concurrent
task, F(1, 23) = 135.56, MSE = .05, p , .001,
η
2
p
= .86, and set size F(1, 23) = 27.36,
MSE = .02, p , .001, η
2
p
= .54. Thus, performance
was highest in the colour condition and lowest in
the binding condition, was disrupted by backward
counting relative to articulatory suppression, and
was better for three-item than for four-item
displays. However, there were no signicant inter-
actions between any of these factors; Stimulus
Condition × Concurrent Task, F(2, 46) = 0.20,
MSE = .03, p = .82, η
2
p
= .01, Stimulus
Condition × Set Size, F(2, 46) = 2.43, MSE = .02,
p = .10, η
2
p
= .10, Concurrent Task × Set Size,
F(1, 23) = 1.71, MSE = .05, p = .20, η
2
p
= .07,
Stimulus Condition × Set Size × Concurrent Task,
F(2, 46) = 0.03, MSE = .02, p = .97, η
2
p
= .01.
An identical pattern of signicance was observed
using d
. There were signicant effects of stimulus
condition, F(2, 46) = 85.88, MSE = .67, p , .001,
η
2
p
= .79, concurrent task, F(1, 23) = 170.52,
MSE = .56, p , .001, η
2
p
= .88, and set size, F(1,
23) = 26.83, MSE = .35, p , .001, η
2
p
= .54.
Again, there were no signicant interactions
between any of these factors; Stimulus Condition ×
Concurrent Task, F(2, 46) = 1.19, MSE = .50,
p = .31, η
2
p
= .05, Stimulus Condition × Set Size,
F(2, 46) = 1.18, MSE = .37, p = .37, η
2
p
= .05,
Concurrent Task × Set Size, F(1, 23) = 3.19,
MSE = .61, p = .09, η
2
p
= .12, Stimulus
Condition × Set Size × Concurrent Task, F(2,
46) = 0.08, MSE = .33, p = .92, η
2
p
= .01.
Finally, A
was also analysed.
1
There were again
signicant effects of stimulus condition, F(2, 46) =
45.85, MSE = .01, p , .001, η
2
p
= .67, concurrent
task, F(1, 23) = 74.70, MSE = .02, p , .001,
η
2
p
= .77, and set size, F(1, 23) = 13.27,
MSE = .01, p = .001, η
2
p
= .37. However, in con-
trast to corrected recognition and d
, there was a
signicant interaction between stimulus condition
and concurrent task, F(2, 46) = 6.01, MSE = .01,
p , .01, η
2
p
= .21. The remaining interactions did
not reach signicance; Stimulus Condition × Set
Size, F(2, 46) = 2.33, MSE = .01, p = .11,
η
2
p
= .09, Concurrent Task × Set Size, F(1, 23) =
0.19, MSE = .02, p = .67, η
2
p
= .01, Stimulus
Condition × Set Size × Concurrent Task, F(2 ,
46) = 0.04, MSE = .01, p = .99, η
2
p
= .01.
In order to identify more clearly the source of the
task by stimulus condition interaction that emerged
using A
, a series of 2 × 2 × 2 analyses were per-
formed, comparing each of the stimulus conditions
in turn. For the colourshape analysis, the key task
by stimulus condition interaction was marginally
signicant, F(1, 23) = 4.26, MSE = .01, p = .051,
η
2
p
= .16. It was also signicant for the colour-
binding analysis, F(1, 23) = 12.11, MSE = .01,
p , .01, η
2
p
= .35. However, the interaction was
not signicant when comparing shape and
binding, F (21, 23) = 2.05, MSE = .01, p = .17,
η
2
p
= .08 (nor was this signicant when examining
set sizes 3 and 4 separately, p = .24 and p = .66,
respectively). Based on Brown and Brockmole
(2010), we also compared the shape and binding
conditions when performed under each concurrent
task condition (collapsing across set size). This
revealed that the binding condition was less accu-
rate than shape during AS, t(23) = 2.53, p , .05,
and during BC, t(23) = 2.60, p , .05, with little
or no increase in this effect between verbal load
conditions.
1
A3× 2 × 2 analysis of the bias measure B
′′
revealed a signicant effect of stimulus condition, F(2, 46) = 6.89, MSE = .21,
p , .01, and a set size by stimulus condition interaction, F(2, 46) = 4.58, MSE = .10, p , .05, thus potentially undermining the
suitability of A
in this context (e.g., Pastore et al., 2003).
6 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0)
ALLEN ET AL.
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 7
Performance accuracy was the primary depen-
dent variable in this study, and indeed accuracy
was emphasized over speed in instructions
to participants. Nevertheless, mean response
latencies (ms) are also provided in Table 1.
Analysis of these data is not reported for the
sake of brevity, but there was no evidence of
any speedaccuracy trade-off patterns concerning
the effects of concurrent task on the different
stimulus conditions.
Discussion
We observed consistent and predicted effects of
stimulus condition and concurrent task in this
experiment, with colour memory substantially
Figure 2. Accuracy (hits false alarms) in Experiment 1 across stimulus conditions, concurrent tasks, and set sizes. AS = articulatory
suppression; BC = backward counting.
Table 1. Accuracy and latency of responses for each stimulus condition, concurrent task load, and set size, in Experiment 1
Set size 3 Set size 4
Measure Task Colour Shape Binding Colour Shape Binding
H FA AS .94 (.02) .75 (.04) .64 (.03) .87 (.03) .60 (.05) .50 (.04)
BC .62 (.05) .41 (.05) .30 (.06) .63 (.05) .33 (.05) .22 (.06)
d
AS 3.67 (.10) 2.60 (.17) 2.07 (.15) 3.27 (.15) 1.95 (.18) 1.55 (.15)
BC 2.15 (.20) 1.32 (.18) 0.92 (.20) 2.13 (.20) 1.02 (.16) 0.66 (.17)
A
AS .98 (.01) .92 (.01) .89 (.01) .96 (.01) .87 (.02) .83 (.02)
BC .88 (.02) .78 (.03) .71 (.04) .88 (.02) .73 (.03) .66 (.04)
Latency (ms) AS 751 (32) 927 (43) 988 (49) 778 (38) 962 (46) 1,006 (55)
BC 1,076 (74) 1,225 (71) 1,210 (49) 1,058 (66) 1,254 (70) 1,311 (82)
Note: Accuracy: H FA (H = hits; FA = false alarms); d
; A
. AS = articulatory suppression; BC = backward counting. Means, with
standard error in parentheses.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0) 7
BINDING AND ATTENTION
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 8
more accurate than shape or binding memory, and a
large disruptiv e effect of attentional load across all
stimulus conditions, thus replicating previous nd-
ings (e.g., Allen et al., 2006). However, the extent
to which these factors interacted varied across the
different ways of measuring memory performance.
Specically, while we observed no stimulus by con-
current task interaction for corrected recognition or
d
, which ts with our previous examination of this
issue, there was a signicant interaction using A
.
Importantly, however, there was no evidence that
this was attributable to larger effects of backward
counting on binding. Instead, it appeared to be
the result of a ceiling effect in the colour/AS con-
dition, where A
was close to its maximum per-
mitted value of 1, which may have articially
limited the potential size of the backward counting
effect in that condition. Overall, across all three
measures of performance, we failed to nd any
clear evidence of a particular role for attention in
shapecolour binding. This would indicate that
the use of longer exposure durations and the con-
tinuation of concurrent task performance through-
out the recognition test phase of each trial are not
key factors in determining whether a greater atten-
tional load effect emerges in binding, relative to
feature memory.
EXPERIMENT 2
A further methodological feature that differs across
studies concerns how stimuli are presented during
the single-feature conditions. In Experiment 1,
and in Allen et al. (2006), the nontested feature
dimension (e.g., shape, during the colour con-
dition) was held constant, meaning that to-be -
remembered displays differed between stimulus
conditions. In contrast, in the work of Brown and
Brockmole (2010; also, e.g., Brockmole, Parra,
Della Sala, & Logie, 2008; Wheeler & Treisman,
2002) displays were constant across conditions,
with shape and colour always varying.
Method of stimulus presentation could impact
on performance in different ways. First, research
suggests that variation in one feature dimension
sometimes helps memory for a second dimension
(e.g., Delvenne & Dent, 2008). As our own pre-
vious research used a neutral, constant feature in
the nontested dimension, this may have rendered
feature conditions more difcult and attention
demanding themselves and thus have concealed a
larger effect of attention on binding. An alternative
possibility is that when both features vary in all con-
ditions (as in Brown & Brockmole, 2010), this
increases the requirement for participants to main-
tain an internal representation of task-relevant
goals, as the display does not serve as an external
cue. Maintaining and applying these goals to task
performance may then become problematic when
an attentional load is added. This would be
expected to affect performance in all conditions,
but might particularly cause the binding condition
to suffer from concurrent task disruption. Either
of these accounts could explain the particular inter-
active effect reported by Brown and Brockmole
(2010).
Therefore, Experiment 2 aimed to replicate the
rst study, but with the exception that features were
allowed to vary in all conditions, resulting in equiv-
alent presentation displays across conditions.
Method
Participants
There were 24 participants (6 males and 18
females, aged between 18 and 22 years, M =
18.75, SD = 1.15) in this experiment. All were
undergraduate students at the University of York,
had English as a rst language, and showed
normal or corrected-to-normal vision.
Materials, design, and procedure
All design features, materials, and testing pro-
cedures from Experiment 1 were implemented
again, with the exception that both shape and
colour feature dimensions varied during presen-
tation in all stimulus conditions. Thus, presen-
tation arrays were equivalent in colour, shape, and
binding conditions, consisting of horizontal rows
of different coloured shapes, and only differed in
terms of the test probe and associated judgement.
Here, procedure followed Expe riment 1, with the
colour condition probed using a colour blob,
8 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0)
ALLEN ET AL.
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 9
shape using a shape outline, and binding using a
coloured shape conjunction.
Results
Mean number of BC responses across conditions
was 3.58 (SE = 0.17), with no signicant main
effects or interactions, p .10. Mean BC error
rate was .23 (SE = .03). There was a marginal
effect of stimulus condition, which was not signi-
cant (after applying GreenhouseGeisser correc-
tion), F(1, 23) = 3.27, MSE = .02, p = .063,
reecting a trend toward a slightly lower error
rate during the shape condition (.19) than during
the colour (.26) or binding (.23) conditions. The
set size and interaction effects were not signicant,
p .26.
As in Experiment 1, recognition judgement
accuracy is reported as corrected recognition (hits
false alarms), d
, and A
. The outcomes of these
responses measures are displayed in Table 2, with
corrected recognition also displayed in Figure 3.
Starting with corrected recognition, we observed
signicant effects of stimulus condition, F(2,
46) = 36.05, MSE = .05, p , .001, η
2
p
= .61, con-
current task, F(1, 23) = 134.99, MSE = .05,
p , .001, η
2
p
= .86, and set size, F(1, 23) = 24.09,
MSE = .02, p , .001, η
2
p
= .51. As in Experiment
1, there was no signicant interaction between
stimulus condition and concurrent task, F(2,
46) = 0.82, MSE = .03, p = .45, η
2
p
= .013, stimu-
lus condition and set size, F(2, 46) = 1.22,
MSE = .03, p = .30, η
2
p
= .05, or concurrent task
and set size, F(1, 23) = 1.71, MSE = .05, p = .20,
η
2
p
= .07, nor the three-way interaction, F(2,
46) = 0.42, MSE = .02, p = .66, η
2
p
= .02. The
one exception was a signicant concurrent task
and set size interaction, F(1, 23) = 7.41,
MSE = .04, p , .05, η
2
p
= .24.
Moving on to d
, there were signi cant effects of
stimulus condition, F(2, 46) = 34.51, MSE = .74,
p , .001, η
2
p
= .60, concurrent task, F(1, 23) =
130.54, MSE = .68, p , .001, η
2
p
= .85, and set
size, F(1, 23) = 26.98, MSE = .40, p , .001,
η
2
p
= .54. In addition, the key interaction between
stimulus condition and concurrent task was signi-
cant, F(2, 46) = 3.43, MSE = .32, p , .05,
η
2
p
= .13. However, examination of the means in
each condition actually indicates a larger effect of
backward counting on colour than on the shape
colour binding condition, a reversal of the result
found by Brown and Brockmole (2010) and also
the A
measure in our Experiment 1. No difference
was observed between the effect of counting on the
binding or shape conditions. The anomalous
colour-based effect probably reects a tendency of
d
to articially inate when performance levels
are high (as in the case of colour-AS). The concur-
rent task by set size interaction was also signicant,
F(1, 23) = 6.18, MSE = .56, p , .05, η
2
p
= .21,
Table 2. Accuracy and latency of responses for each stimulus condition, concurrent task load, and set size, in Experiment 2
Set size 3 Set size 4
Measure Task Colour Shape Binding Colour Shape Binding
HFA AS .86 (.03) .73 (.04) .64 (.04) .71 (.03) .55 (.04) .45 (.06)
BC .50 (.03) .35 (.06) .28 (.05) .50 (.04) .28 (.04) .26 (.04)
d
AS 3.19 (.15) 2.52 (.16) 2.06 (.16) 2.73 (.16) 1.81 (.17) 1.41 (.23)
BC 1.66 (.23) 1.20 (.21) 0.93 (.19) 1.61 (.15) 0.92 (.15) 0.75 (.14)
A
AS .96 (.01) .92 (.01) .88 (.02) .93 (.01) .85 (.02) .79 (.03)
BC .82 (.04) .75 (.03) .71 (.03) .84 (.02) .71 (.03) .69 (.03)
Latency (ms) AS 762 (36) 945 (36) 940 (35) 778 (38) 962 (46) 1,006 (55)
BC 1,259 (90) 1,352 (94) 1,382 (89) 1,301 (83) 1,494 (124) 1,299 (98)
Note: Accuracy: H FA (H = hits; FA = false alarms); d
; A
. AS = articulatory suppression; BC = backward counting. Means, with
standard error in parentheses.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0) 9
BINDING AND ATTENTION
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 10
while the Stimulus Condition × Set Size, F(2,
46) = 0.71, MSE = .52, p = .50, η
2
p
= .03, and
Stimulus Condition × Set Size × Concurrent
Task interactions, F(2, 46) = 0.02, MSE = .37,
p = .98, η
2
p
= .01, were not.
Finally, analysis of A
2
revealed signicant effects
of stimulus condition, F(2, 46) = 5.82, MSE = .86,
p , .05, η
2
p
= .20, concurrent task, F(1, 23) =
106.23, MSE = .45, p , .001, η
2
p
= .82, and set
size, F(1, 23) = 30.66, MSE = .032, p , .001,
η
2
p
= .57. However, there were no signicant inter-
actions between any of these factors; Stimulus
Condition × Concurrent Task, F(2, 46) = 1.84,
MSE = .29, p = .19, η
2
p
= .07, Stimulus
Condition × Set Size, F(2, 46) = 0.13, MSE = .56,
p = .72, η
2
p
= .01, Concurrent Task × Set Size, F
(1, 23) = 3.85, MSE = .64, p = .06, η
2
p
= .14,
Stimulus Condition × Set Size × Concurrent Task,
F(2, 46) = 0.02, MSE = .36, p = .89, η
2
p
= .01.
As in Experiment 1, analysis of response
latencies (provided in Table 2) produced no
evidence of any speedaccuracy trade-offs concern-
ing the impact of concurrent load on binding versus
single-feature conditions.
Discussion
Using a methodology in which equivalent displays
were encountered in each stimulus condition, we
again found no evidence of a larger role for execu-
tive attention in feature binding relative to memory
for the individual features. Thus, across all three
measures of performance, the impact of backward
counting was at least as large for the colour and
shape conditions as it was for binding memory.
The results also provided no support for the idea
that maintaining task goals was more attention
demanding when stimulus displays were equivalent
in the shape, colour, and binding conditions. Thus
the amount of interference from backward counting
was virtually the same as that in Experiment 1, in
Figure 3. Accuracy (hits false alarms) in Experiment 2 across stimulus conditions, concurrent tasks, and set sizes. AS = articulatory
suppression; BC = backward counting.
2
As in Experiment 1, analysis of B
′′
revealed a signicant effect of stimulus condition, F(2, 46) = 12.82, MSE = .18, p , .001.
10 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0)
ALLEN ET AL.
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 11
which stimulus type gave an external cue to the task
goal on any trial. Backward counting did, however,
tend to disrupt memory more when set size was
lower, with all three measures showing this inter-
action (though in the case of A
it was marginally
nonsignicant). This interaction was unexpected
and has no obvious theoretical interpretation.
Looking back, we note there was a nonsignicant
trend for a corresponding interaction in
Experiment 1.
GENERAL DISCUSSION
To recap, the present experiments investigated
whether retaining shapecolour bindings in
working memory is more demanding of execu-
tive-based attention than retaining individual fea-
tures (Baddeley, 2000; Wheeler & Treisman,
2002), or whether there is no difference (Allen
et al., 2006; Baddeley et al., 2011). Followin g pre-
vious research, this question was addressed using
dual-task methodology to see whether an irrelevant
concurrent attentional load disrupted memory for
feature bindings more than memory for features,
or to the same extent. Our own earlier experiments
of this sort have consistently supported the view
that remembering feature bindings is no more
demanding of attention than remembering individ-
ual features (Allen et al., 2006). However, a similar
experiment by Brown and Brockmole (2010)
reported conicting results favouring the alternative
hypothesis. The specic aim of the present study
was to examine possible methodological reasons
for the discrepant results in order to resolve the
controversy.
Although the two lines of experimentation were
conceptually similar, they differed on a number of
methodological details, any of which might have
contributed to the different outcomes (as noted
by Brown & Brockmole, 2010). Among the differ-
ences we considered were (a) whether the concur-
rent task overlapped with the decision phase of
the recognition memory task (this was the case in
Brown & Brockmole, 2010, but not in Allen
et al., 2006), (b) stimulus exposure duration
(Brown & Brockmole, 2010, used 900 ms
whereas Allen et al., 2006, used 250 ms), (c)
memory set size (Brown & Brockmole, 2010,
used displays of three items where as Allen et al.,
2006, used four), (d) implementatio n of single-
feature conditions (Brown & Brockmole, 2010,
used the same stimuli in their feature and binding
conditions whereas Allen et al., 2006, held the irre-
levant feature constant in their feature conditions),
and nally (e) the metric for recognition memory
performance (Brown & Brockmole, 2010, used A
whereas Allen et al., 2006, used corrected recog-
nition, i.e., hits false alarms).
In the present experiments, the concurrent task
continued throughout the decision phase of the
memory task, stimuli were presented for 1,000
ms, memory set size was either 3 or 4 items,
and memory performance was analysed in terms
of A
, corrected recognition, and d
. Over and
above this, stimuli in the feature conditions
varied on a single feature in Experiment 1 (as in
Allen et al., 2006), but on both relevant and irre-
levant features in Experiment 2 (as in Brown &
Brockmole, 2010).
At a general level, the results can be summarized
quite simply. Thus in both experiments, and for all
measures, a concurrent attention-demanding task
impaired memory in both the single-feature and
the binding conditions. Importantly, there was no
consistent or compelling evidence across exper-
iments and measures to suggest that concurrent
attentional load disrupted memory for feature bind-
ings to a greater extent than memory for features. In
the only instance where there was a hint of such an
interaction (in Experiment 1), it was found for
colour but not for shape and for A
but not for d
or corrected recognitio n. Moreover, there were
grounds for attributing this isolated interaction to
a ceiling effect in the condition testing memory
for colour under articulatory suppression. Given
also the possibility that Brown and Brockmoles
(2010) interaction may have been due to a similar
ceiling effect, we take the present results as consist-
ent with previous evidence that encoding, main-
taining, and retrieving feature bindings in visual
working memory does not require more attentional
support than do individual features.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0) 11
BINDING AND ATTENTION
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 12
As explained earlier, resolution of apparently
contradictory results of similar dual-task studies is
important for the debate between theoretical
accounts in which attention plays a special role in
encoding, maintaining, and retrieving feature bind-
ings in visual working memory (Baddeley, 2000;
Wheeler & Treisman, 2002) and alternative
accounts that assume that feature binding is
largely automatic (Allen et al., 2006; Baddeley
et al., 2011; Luck & Vogel, 1997). Given the
outcome of the present experiments, we remain
condent in holding to the latter position. Indeed
we regard the present results as consistent with a
recent modication of initial ideas about the role
of the episodic buffer in binding proposed by
Baddeley (2000). In our revised account, we
assume that binding takes place automatically
within the visuospatial sketchpad, which we
regard as a multilevel store holding information
about both feature and object levels (Baddeley
et al., 2011). The sketchpad feeds information
about objects into the episodic buffer where it is
available for conscious control and manipulation.
In our revised account, information about feature
bindings differs from information about single
features in terms of its fragility, rather than its
demands for executive resources. In other
studies, we have found that binding information
is particularly susceptible to overwriting from sub-
sequent visual inputs (Allen et al., 2006; se e also
Brown & Brockmole, 2010) with the amount of
overwriting determined to some extent by
whether subsequently encountered information
contains one or more plausible single features
(Ueno, Allen, et al., 2011; Ueno, Mate, et al.,
2011; though see Woodman & Vogel, 2008, for
evidence indicating the possibility of selective
encoding of task-relevant information). This
fragility may be the core factor underlying other
forms of evidence demonstrating loss of binding
information during retention (e.g., Brown &
Brockmole, 2010; Fougnie & Marois, 2009;
Wheeler & Treisman, 2002), although the
precise mechanisms, including the role of atten-
tion, involved in maintaining feature bindings in
working memory and determining their robustness
or fragility are yet to be fully mapped out.
We accept of course that the present experimen-
tal data are consistent with other explanations, but
offer the above account as embracing a wider set of
ndings and a broad framework for asking further
questions about binding, both within and beyond
the visual domain (see Baddeley et al., 2011).
While our ndings t with the assumption of rela-
tively automatized binding processes at encoding,
they contrast with the claims of vanRullen (2009)
that the initial on-demand processing of arbitrary
feature bindings is particularly reliant on attention.
It should, however, be noted that the arguments
put forward by vanRu llen draw on object recog-
nition tasks rather than measures of working
memory and are concerned more with spatial atten-
tion, as opposed to the general executive-based
attentional resources currently under examination.
Furthermore, vanRullens claims are generally based
on binding within feature dimensions (e.g., colour
colour, or orientationorientation). A number of
studies have indicated that within-dimension
binding may not operate in the same way as
binding between dimensions such as colour and
shape (e.g., Delvenne & Bruyer, 2004; Parra,
Abrahams, Logie, & Della Sala, 2008; Wheeler &
Treisman, 2002). Indeed, further research would be
required to examine whether attentional support is
important for within-dimension binding in
working memory. Instead, our ndings are more in
line with the claims made by Hommel and Colzato
(2009). Like vanRullen (2009), they also distin-
guished between new and arbitrary bindings and
familiar conjunctions that already exist in long-term
memory, but argued that both are encoded automati-
cally in working memory.
Finally, one incidental aspect of the present
results deserves special mention. This is the obser-
vation of inconsistencies between experimental
outcomes depending on the way recognition
memory performance is measured. Although A
,
corrected recognition, and d
were consistent with
respect to the main effects of experimental manip-
ulations, they were inconsistent at the ner grain of
interactions, most notably that between concurrent
task and stimulus condition. Given that this inter-
action was the central issue here, we are led to con-
sider broader methodological implications of the
12 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0)
ALLEN ET AL.
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 13
present ndings. First, we note that avoiding
ceiling and oor effects presents a pervasive chal-
lenge in research on visual working memory. The
capacity of visual working memory is sharply
limited to just three or four objects (Luck &
Vogel, 1997), giving limited scope for adjusting
difculty levels by varying set size. A second impli-
cation is that we should take seriously cautions
regarding the measurement of recognition
memory performance (Pastore et al., 2003). We
note particularly Pastore et al.s (2003) observation
of large discrepancies between A
and d
at high
levels of performance, such that A
underestimates
performance relative to d
unless the decision cri-
terion is unbalanced. The present results indicate
that these caveats concerning ceiling effects and
their consequences for different measures of recog-
nition memory performance can be very important
indeed when we are interested in identifying
interactions. In particular they seem likely to have
contributed to the inconsistent patterns of signi -
cant interaction between concurrent task and
stimulus condition for A
in Experiment 1, d
in
Experiment 2, and A
in Brown and Brockmole
(2010). It is only through exploration of the discre-
pancy between our ndings and those of Brown
and Brockmole that the import ance of these
factors has emerged. We suggest that future exper-
iments should take account of such methodological
issues to ensure robustness and replicability of nd-
ings and thus help speed progress towards the goal
of establishing a cumulative body of empirical evi-
dence on the role of attention in feature binding
in visual working memory.
Original manuscript received 6 December 2011
Accepted revision received 15 March 2012
First published online 6 June 2012
REFERENCES
Allen, R. J., Baddeley, A. D., & Hitch, G. J. (2006). Is
the binding of visual features in working memory
resource-demanding? Journal of Experimental
Psychology: General, 135, 298313.
Allen, R. J., Hitch, G. J., & Baddeley, A. D. (2009).
Cross-modal binding and working memory. Visual
Cognition, 17, 83102.
Alvarez & Cavanagh, 2009
Alvarez, G. A., & Thompson, T. W. (2009).
Overwriting and rebinding: Why feature-switch
detection tasks underestimate the binding capacity
of visual working memory. Visual Cognition, 17,
141159.
Awh, E., Barton, B., & Vogel, E. K. (2007). Visual
working memory represents a xed number of items
regardless of complexity. Psychological Science, 18(7),
622628.
Baddeley, A. D. (2000). The episodic buffer: A new
component of working memory? Trends in Cognitive
Sciences, 4(11), 417423.
Baddeley, A. D., Allen, R. J., & Hitch, G. J. (2011).
Binding in visual working memory: The role of the
episodic buffer. Neuropsychologia, 49, 13931400.
Baddeley, A. D., & Hitch, G. J. (1974). Working
memory. In G.A. Bower (Ed.), The psychology
of learning and motivation: Advances in research
and theory (Vol. 8, pp. 4789). New York, NY:
Academic Press.
Brockmole, J. R., Parra, M. A., Della Sala, S., & Logie,
R. H. (2008). Do binding decits account for
age-related decline in visual working memory?
Psychonomic Bulletin & Review, 15, 543547.
Brown, L. A., & Brockmole, J. R. (2010). The role of
attention in binding visual features in working
memory: Evidence from cognitive ageing. The
Quarterly Journal of Experimental Psychology, 63,
20672079.
DellAcqua, R., & Jolicoeur, P. (2000). Visual encoding
of patterns is subject to dual-task interference.
Memory & Cognition, 28, 184191.
Delvenne, J.-F., & Bruyer, R. (2004). Does visual short-
term memory store bound features? Visual Cognition,
11, 127.
Delvenne, J.-F., Cleeremans, A., & Laloyaux, C. (2010).
Feature bindings are maintained in visual short-term
memory without sustained focused attention.
Experimental Psychology, 57(2), 108116.
Delvenne, J.-F., & Dent, K. (2008). Distinctive shapes
benet short-term memory for associations, not
colors. Perception & Psychophysics, 70(6), 10241031.
Duncan, J. (1984). Selective attention and the organiz-
ation of visual information. Journal of Experimental
Psychology: General, 113(4), 501517.
Elsley, J. V., & Parmentier, F. B. R. (2009). Is verbal
spatial binding in working memory impaired by a
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0) 13
BINDING AND ATTENTION
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 14
concurrent memory load? The Quarterly Journal of
Experimental Psychology, 62(9), 16961705.
Fougnie, D., Asplund, C. L., & Marois, R. (2010).
What are the units of storage in visual working
memory? Journal of Vision, 10(12), 2711.
Fougnie, D., & Marois, R. (2009). Attentive tracking
disrupts feature binding in visual working memory.
Visual Cognition, 17, 4866.
Gajewski, D. A., & Brockmole, J. R. (2006). Feature
bindings endure without attention: Evidence from
an explicit recall task. Psychonomic Bulletin &
Review, 13, 581587.
Gorgoraptis, N., Catalao, R. F. G., Bays, P. M., &
Husain, M. (2011). Dynamic updating or working
memory resources for visual objects. The Journal of
Neuroscience, 31(23), 85028511.
Hommel, B., & Colzato, L. S. (2009). When an object is
more than a binding of its features: Evidence for two
mechanisms of visual feature integration. Visual
Cognition, 17(1/2), 120140.
Hyun, J., Woodman, G. F., Vogel, E. K., Hollingworth,
A., & Luck, S. J. (2009). The comparison of visual
working memory representations with perceptual
inputs. Journal of Experimental Psychology: Human
Perception and Performance, 35(4), 11401160.
Johnson, J. S., Hollingworth, A., & Luck, S. J. (2008).
The role of attention in the maintenance of feature
bindings in visual short-term memory. Journal of
Experimental Psychology: Human Perception &
Performance, 34, 4155.
Karlsen, P. J., Allen, R. J., Baddeley, A. D., & Hitch, G.
J. (2010). Binding across space and time in visual
working memory. Memory and Cognition, 38,
292303.
Logie, R. H., Brockmole, J. R., & Vandenbroucke,
A. R. E. (2009). Bound feature combinations in
visual short-term memory are fragile but in uence
long-term learning. Visual Cognition, 17, 160179.
Luck, S. J., & Vogel, E. K. (1997). The capacity of visual
working memory for features and conjunctions.
Nature, 390, 279281.
McNicol, D. (1972). A primer of signal detection theory.
London, UK: George Allen & Unwin.
Mitchell, K. J., Johnson, M. K., Raye, C. L., Mather,
M., & D
Esposito, M. (2000). Aging and reective
processes of working memory: binding and test load
decits. Psychology and Aging, 15(3), 527541.
Morey, C. C. (2011). Maintaining binding in working
memory: Comparing the effects of intentional
goals and incidental affordances. Consciousness and
Cognition, 20, 920927.
Morey, C. C., & Cowan, N. (2004). When visual and
verbal memories compete: Evidence of cross-
domain limits in working memory. Psychonomic
Bulletin and Review, 11, 296301.
Morey, C. C., & Cowan, N. (2005). When do visual and
verbal memory conict? The importance of working
memory load and retrieval. Journal of Experimental
Psychology: Learning, Memory & Cognition, 31,
703713.
Olson, I., & Jiang, Y. (2002). Is visual short-term
memory object-based? Perception & Psychophysics,
64, 10551067.
Parra, M. A., Abrahams, S., Logie, R. H., & Della Sala,
S. (2008). Age and binding within-dimension fea-
tures in visual short-term memory. Neuroscience
Letters, 449, 15.
Parra, M. A., Della Sala, S., Logie, R. H., & Abrahams,
S. (2009). Selective impairment in visual short-term
memory binding. Cognitive Neuropsychology, 26(7),
583605.
Pastore, R. E., Crawley, E. J., Berens, M. S., & Skelly,
M. A. (2003). Nonparametric A
and other
modern misconceptions about signal detection
theory. Psychonomic Bulletin & Review, 10(3),
556569.
Pollack, I., & Norman, D. A. (1964). A non-parametric
analysis of recognition memory. Psychonomic Science,
1, 125126.
Snodgrass, J. G., & Corwin, J. (1988). Pragmatics of
measuring recognition memory: Applications to
dementia and amnesia. Journal of Experimental
Psychology: General, 117, 3450.
Song, J-H., & Jiang, Y. (2006). Visual working memory
for simple and complex features: An fMRI study.
NeuroImage, 30, 963972.
Stanislaw, H., & Todorov, N. (1999). Calculation of
signal detection theory measures. Behavior Research
Methods, Instruments, & Computers, 31, 137149.
Stevanovski, B., & Jolicoeur, P. (2007). Visual short-
term memory: Central capacity limitations in short-
term consolidation. Visual Cognition, 15, 532563.
Ueno, T., Allen, R. J., Baddeley, A. D., Hitch, G. J., &
Saito, S. (2011). Disruption of visual feature binding
in working memory. Memory & Cognition, 39, 1223.
Ueno, T., Mate, J., Allen, R. J., Hitch, G. J., & Baddeley,
A. D. (2011). What goes through the gate? Exploring
interference with visual feature binding.
Neuropsychologia, 49, 15971604.
vanRullen, R. (2009). Binding hardwired versus on-
demand feature conjunctions. Visual Cognition, 17
(1/2), 103119.
14 THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0)
ALLEN ET AL.
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 15
Vogel, E. K., Woodman, G. F., & Luck, S. J. (2001).
Storage of features, conjunctions, and objects in
visual working memory. Journal of Experimental
Psychology: Human Perception and Performance, 27
(1), 92114.
Wheeler, M. E., & Treisman, A. M. (2002). Binding in
short-term visual memory. Journal of Experimental
Psychology: General, 131, 4864.
Woodman, G. F., & Vogel, E. K. (2008). Selective
storage and maintenance of an objects features in
visual working memory. Psychonomic Bulletin &
Review, 15(1), 223229.
Xu, Y. (2002). Encoding color and shape from
different parts of an object in visual short-term
memory. Perception & Psychophysics, 64(8),
12601280 .
Xu, Y. (2006). Understanding the object benet in visual
short-term memory: The roles of feature proximity
and connectedness. Perception & Psychophysics, 68
(5), 815828.
Yeh, Y. Y., Yang, C. T., & Chiu, Y. C. (2005). Binding
or prioritization: The role of selective attention in
visual short-term memory. Visual Cognition, 12(5),
759799.
THE QUARTERLY JOURNAL OF EXPERIMENTAL PSYCHOLOGY, 2012, 00 (0) 15
BINDING AND ATTENTION
Downloaded by [University of Leeds] at 04:39 23 June 2012
Page 16
  • Source
    • "There is no doubt that the inclusion of the episodic buffer improved the scope of the model. Nevertheless, it did not result in a reduction of the power of the central executive because further research has shown that the hypothesized binding operations attributed to the episodic buffer do not involve executive control (Allen, Baddeley, & Hitch, 2006; Allen, Hitch, Mate, & Baddeley, 2012; Baddeley et al., 2010; Baddeley, Hitch, & Allen, 2009 ). In another effort to fractionate the central executive, Miyake et al. (2000) proposed replacing it with specific executive functions, such as changing focus from one intention to another (set shifting); adapting the contents of WM by adding, replacing , or deleting contents (memory updating); and suppressing or decreasing the degree of activation of WM contents (inhibition). "
    Dataset: Full text
    Full-text · Dataset · Feb 2016
  • Source
    • "Subsequent studies have found that secondary tasks engaging spatial attention produce no specific decrement in binding memory, with robust memory for binding in the absence of sustained visual attention (Delvenne, Cleeremans, & Laloyaux, 2010; Gajewski & Brockmole, 2006; Johnson et al., 2008; Shen, Huang, & Gao, 2015; van Lamsweerde & Beck, 2012). Complementary evidence has been observed in studies focusing on executive attention (Allen, Baddeley, & Hitch, 2006; Allen, Hitch, Mate, & Baddeley, 2012; Morey & Bieler, 2013; van Lamsweerde, Beck, & Elliott, 2015). Recently, Shen et al. (2015 showed that several object-based attention tasks selectively impair binding in VWM, but this is to be expected given that "
    [Show abstract] [Hide abstract] ABSTRACT: There is substantial debate over whether visual working memory (VWM) and visual attention constitute a single system for the selection of task-relevant perceptual information or whether they are distinct systems that can be dissociated when their representational demands diverge. In the present study, we focused on the relationship between visual attention and the encoding of objects into VWM. Participants performed a color change-detection task. During the retention interval, a secondary object, irrelevant to the memory task, was presented. Participants were instructed either to execute an overt shift of gaze to this object (Experiments 1-3) or to attend it covertly (Experiments 4 and 5). Our goal was to determine whether these overt and covert shifts of attention disrupted the information held in VWM. We hypothesized that saccades, which typically introduce a memorial demand to bridge perceptual disruption, would lead to automatic encoding of the secondary object. However, purely covert shifts of attention, which introduce no such demand, would not result in automatic memory encoding. The results supported these predictions. Saccades to the secondary object produced substantial interference with VWM performance, but covert shifts of attention to this object produced no interference with VWM performance. These results challenge prevailing theories that consider attention and VWM to reflect a common mechanism. In addition, they indicate that the relationship between attention and VWM is dependent on the memorial demands of the orienting behavior. (PsycINFO Database Record
    Full-text · Article · Feb 2016 · Journal of Experimental Psychology Human Perception & Performance
  • Source
    • "However, we would note that concurrent backward counting impacted on all outcome measures in Experiment 1 (recall of action pairs, along with movements and binding of movement to object features). This would indicate that the central executive likely plays a general role in supporting all elements of performance, rather than specifically supporting feature binding (Allen et al., 2006; Allen, Hitch, Mate, & Baddeley, 2012; Baddeley et al., 2011). Interfering with spatial coding by concurrent spatial tapping during encoding led to a relatively small and non-significant recall decrement, whereas when visuospatial cues were completely blocked via eye-closure, the performance of both verbal and enacted recall dropped significantly. "
    [Show abstract] [Hide abstract] ABSTRACT: This study investigated the involvement of working memory (WM) in following spoken instructions using concurrent tasks designed to disrupt components of the Baddeley and Hitch WM model [Baddeley, A. D., & Hitch, G. J. (1974). Working memory. In G. H. Bower (Ed.), The psychology of learning and motivation: Advances in research and theory (Vol. 8, pp. 47–89). New York, NY: Academic Press]. In each of three experiments, participants were presented with sequences of instructions to be either verbally repeated or physically performed using relevant objects. Backward counting, articulatory suppression, and eye closure during instruction encoding all disrupted recall, and also impaired recall of the linkage between movements and objects. Recall of actions was more accurate when they were physically enacted than repeated verbally, an advantage that was not affected by concurrent tasks. These findings indicate that aspects of the recall of spoken instructions including the binding of constituent movements to objects draw on multiple WM resources. The benefits of physical enactment of the instructed sequence do not appear to depend on the components of WM investigated in these studies.
    Full-text · Article · Oct 2015 · Journal of Cognitive Psychology
Show more