Attentional bias, distractibility and short-term memory in anxiety

Abstract

Abstract Cognitive effects of anxiety have been amply documented. Anxiety has been linked with an attentional bias toward threat, distractibility, and reductions in short-term memory (STM) capacity. These three functions have rarely been investigated jointly and permeability may account for some of the effects documented. In this experiment, we examine these three cognitive functions using one verbal and one visuospatial task. In the irrelevant speech paradigm, participants had to remember strings of letters while irrelevant neutral or threatening speech was presented. In the visuospatial sandwich paradigm, participants were asked to remember sequences of visuospatial targets sometimes presented within irrelevant distracters. We examined the links between state anxiety, worry, and indices of attentional bias toward threat, distractibility from neutral stimuli, and STM capacity. Results show that state anxiety was uniquely linked with impairments in STM while worry was more particularly related to distractibility, independently from permeability between the different cognitive functions. Attentional bias toward threat was linked with variance common to both anxiety and worry. An examination of clinical and non-clinical subgroups suggests that subjective threat perception and attentional bias toward threat are features that are particularly characteristic of clinical levels of anxiety. Our findings confirm the important links between anxiety and basic cognitive functions.

Full text

Attentional bias, distractibility and short-term memory in anxiety
Marie-Laure B. Lapointe
a,c
, Isabelle Blanchette
b
,Me´lanie Duclos
a
,
Fre´de´ric Langlois
b
*, Martin D. Provencher
a
and Se´bastien Tremblay
a
a
E
´cole de Psychologie, Universite´ Laval, Que´bec, Canada;
b
De´partement de Psychologie,
Universite´ du Que´bec a` Trois-Rivie`res, Que´ bec, Canada;
c
Institut Universitaire en sante´ mentale
de Que´bec, Que´bec, Canada
(Received 6 July 2011; final version received 19 April 2012)
Cognitive effects of anxiety have been amply documented. Anxiety has been
linked with an attentional bias toward threat, distractibility, and reductions
in short-term memory (STM) capacity. These three functions have rarely been
investigated jointly and permeability may account for some of the effects
documented. In this experiment, we examine these three cognitive functions
using one verbal and one visuospatial task. In the irrelevant speech paradigm,
participants had to remember strings of letters while irrelevant neutral or
threatening speech was presented. In the visuospatial sandwich paradigm,
participants were asked to remember sequences of visuospatial targets sometimes
presented within irrelevant distracters. We examined the links between state
anxiety, worry, and indices of attentional bias toward threat, distractibility from
neutral stimuli, and STM capacity. Results show that state anxiety was uniquely
linked with impairments in STM while worry was more particularly related to
distractibility, independently from permeability between the different cognitive
functions. Attentional bias toward threat was linked with variance common to
both anxiety and worry. An examination of clinical and non-clinical subgroups
suggests that subjective threat perception and attentional bias toward threat are
features that are particularly characteristic of clinical levels of anxiety. Our
findings confirm the important links between anxiety and basic cognitive
functions.
Keywords: anxiety; attentional bias; distractibility; short-term memory; threat
perception
Introduction
A large amount of research has documented the important ways in which cognitive
functions are altered or impaired as a function of anxiety. According to some
authors, cognitive alterations may even play a key role as a causal and maintenance
factor for anxiety disorders (Mathews & MacLeod, 2002; Van den Hout, Tenney,
Huygens, & Merckelbach, 1995; Williams, Mathews, & MacLeod, 1996). Anxiety has
been linked with specific cognitive effects, including an attentional bias toward
threatening information (e.g., Fox, 1996; MacLeod, Mathews, & Tata, 1986), greater
susceptibility to distraction (e.g., Fox, 1993; Pallack, Pittman, Heller, & Munson,
1975), and possible alterations in short-term memory (STM) (Derakshan & Eysenck,
1998; Hayes, Hirsch, & Mathews, 2008). These first stages of information processing
*Corresponding author. Email: fre´de´ric.langlois@uqtr.ca
Anxiety, Stress, & Coping, 2013
Vol. 26, No. 3, 293313, http://dx.doi.org/10.1080/10615806.2012.687722
#2013 Taylor & Francis

are important intrinsically and furthermore can have cascade effects on subsequent
higher level cognitive functions. Though much research has examined each of these
three functions independently, they have rarely been examined jointly. Permeability
between these cognitive tasks may account for some of the findings in the literature.
The main goal of this article is to investigate the possible links between these three
cognitive domains as they relate to anxiety. An additional goal is to examine whether
anxiety differentially impacts verbal and visuospatial information processing.
Mialet (2000) suggests that anxiety is characterized by three main cognitive
dysfunctions in attentional and memory domains: (1) attentional bias toward
information perceived as threatening, (2) greater susceptibility to distraction, and (3)
reduction in STM capacity. The attentional perturbation that has been most widely
studied in anxiety is the attentional bias toward threatening information (see
Mathews & MacLeod, 2005, for a review). The attentional bias toward threatening
stimuli in high trait anxiety has been well established in the literature using a range of
paradigms (e.g., the emotional Stroop task, Fox, 1993; Richards & French, 1990;
Richards, French, Johnson, & Naparstek, 1992; Van den Hout et al., 1995; the dot
probe task, Bradley, Mogg, Falla, & Hamilton, 1998; MacLeod & Mathews, 1988;
Mogg, Mathews, Bird, & Macgregor-Morris, 1990; Mogg et al., 2000; and the spatial
cueing paradigm, Fox, Russo, Bowles, & Dutton, 2001; Georgiou et al., 2005; Yiend
& Mathews, 2001). Ample empirical evidence has shown that anxious individuals
tend to give processing priority to information perceived as threatening. This has
been confirmed in a recent meta-analysis (Bar-Haim, Lamy, Pergamin, Bakermans-
Kranenburg, & van Ijzendoorn, 2007).
In parallel to the attentional bias toward threat, empirical findings suggest that
high-anxious individuals may also present a general vulnerability to distraction, even
in relation to neutral stimuli. This phenomenon, distractibility, has been defined as
the difficulty to ignore task-irrelevant stimuli (Mialet, 2000). Several studies using
paradigms such as the classic Stroop task (e.g., Fox, 1993; Pallack et al., 1975), letter
transformation task (e.g., Eysenck & Graydon, 1989), and selective search paradigm
(e.g., Mathews, May, Mogg, & Eysenck, 1990) have demonstrated that anxious
participants exhibit higher sensitivity to distraction than non-anxious subjects.
There is a need to better distinguish between the attentional bias toward threaten-
ing stimuli and distractibility. Several studies of distractibility have presented
some conceptual problems and it remains unclear whether these studies included
entirely neutral materials or if, instead, they may have included stimuli that were
somewhat threatening. Thus, it is possible that some of the effects interpreted as
increased distractibility have actually been the result of attentional bias toward threat
(see Eysenck & Byrne, 1992). Indeed, few studies on distractibility have included
manipulation checks to ensure that the stimuli were in fact neutral, especially to high-
anxious individuals. In addition, most of this research focusing on distractibility was
conducted during the 1980s, when the assessment of anxiety was sometimes imprecise.
At that time, the definition of general anxiety disorder (GAD) was still unclear and the
focus of much debate (see Wittchen & Hoyer, 2001). Thus it is important to establish
whether differences in anxiety levels are associated with increased distractibility from
neutral stimuli, with adequate measures of both anxiety and of the emotional nature of
the stimuli.
While attentional bias toward threat may have confounded some of the
conclusions related to distractibility, general distractibility could also explain some
294 M.-L.B. Lapointe et al.

of the variance in attentional bias toward threat. If high-anxious participants are
more susceptible to distraction, it would be more difficult to inhibit attentional
capture by irrelevant threat stimuli. A general propensity to distractibility may
simply increase the likelihood that task-irrelevant stimuli will attract attention. This
is even more important given that the attentional bias toward threat is not necessarily
restricted to high-anxious individuals. Allocation of attention to threat, though more
pronounced in high-anxious or clinically anxious individuals, has also been observed
in unselected samples (Blanchette, 2006; Fox, Griggs, & Mouchlianitis, 2007). There
is evidence, for example, that low-anxious individuals allocate their attention toward
threat under certain conditions when the intensity of stimuli reaches a certain
threshold (Yiend & Mathews, 2001). Another robust finding is that high-anxious
individuals evaluate stimuli as more threatening than non-anxious individuals
(Blanchette & Richards, 2010). In other words, what is neutral for non-anxious
individuals may be threatening for high-anxious individuals. What is interpreted as
an attentional bias toward threat in high-anxious individuals may actually result
from the combination of higher distractibility and increased threat evaluation. Thus,
there is a need to adequately differentiate between a general propensity to
distractibility and a specific allocation of attentional resources toward threatening
stimuli, accounting for the fact that the same stimuli may not be perceived to be as
neutral by anxious and non-anxious individuals.
A further cognitive domain that may be linked with attention and distractibility is
STM. The characterization of memory functioning in anxiety has been less
investigated than attentional perturbations (Mialet, 2000). Most research has studied
memory dysfunctions in order to highlight a memory bias toward threatening
material (e.g., Bradley, Mogg, & Williams, 1994; Reidy & Richard, 1997), which is
consistent with the fact that an attentional bias toward threat would affect encoding
and thus later memory performance. The few studies that have focused on STM for
neutral stimuli in anxiety have indicated that alterations in STM may also
characterize the cognitive profile of anxious individuals (Eysenck, 1979; Eysenck,
Payne, & Derakshan, 2005; Hayes et al., 2008). It has been demonstrated with
different paradigms that high state anxiety reduces performance in verbal STM (e.g.,
digit span task, Darke, 1988; Firetto & Davey, 1971; Hodges & Spielberger, 1969;
reading span task, Sorg & Whitney, 1992). Elevated trait anxiety has also been
associated with lower verbal STM performance (Darke, 1988). It is not clear whether
attentional processes, such as distractibility, are responsible for memory impairments
or whether anxiety is characterized by a genuine reduction in STM capacity.
Increased distractibility, resulting in a reduced ability to inhibit task irrelevant
information, could account for lower performance on complex working memory
measures (such as the reading span task). To better characterize memory
performance, stimuli devoid of affective value must be used in order to determine
whether deleterious effects of anxiety result from a genuine reduction in STM span
or increased distractibility.
We have so far considered all forms of anxiety indiscriminately. We have reviewed
experiments documenting effects linked with transient levels of state anxiety,
differences attributable to stable individual differences such as trait anxiety, and
the cognitive profiles of clinically anxious individuals suffering from generalized
anxiety disorders. Though there may be many commonalities between the effects of
these different forms of anxiety, there is also a need to examine more specifically
Anxiety, Stress, & Coping 295

whether effects observed with non-pathological levels of anxiety also characterize the
profile of clinically anxious populations, and whether distinct cognitive profiles are
linked with pathological and non-pathological anxiety.
A further issue of interest is whether the cognitive alterations related to anxiety
are similarly evident in the verbal and visuospatial domains. One of the main
constituent features of pathological anxiety is the tendency to worry. Borkovec,
Robinson, Pruzinsky and DePree (1983) described worry as a ‘‘chain of thoughts and
images, negatively affect-laden and relatively uncontrollable.’’ The worry process can
also be considered as an attempt to engage in mental problem solving of real issues
whose outcome is uncertain but includes the possibility of one or more negative
outcomes (Borkovec et al., 1983). Characteristically, worry would be expected to
alter verbal information processing (see Leigh & Hirsch 2011 for related empirical
evidence). The attentional bias toward threat has been robustly established with both
verbal and visual stimuli, using for example the Stroop and dot-probe tasks,
respectively. Examinations of STM have, however, mostly been examined in the
verbal domain, though some studies have shown that a high tendency to worry
(Crowe, Matthews, & Walkenhorst, 2007) and elevated state anxiety (Shackman
et al., 2006) are associated with lower performance on the Corsi blocks task (Corsi,
1973), a task assessing visuospatial STM. Distractibility has been mainly examined
with verbal materials and may logically be expected to relate more to worry than to
anxiety generally. Thus, an additional focus of this article is to characterize whether
processing of visual and verbal information is similarly altered as a result of anxiety,
especially in relation to distractibility and STM capacity.
Methodological approach
To summarize, studies converge on the conclusion that an attentional bias toward
threatening material, greater susceptibility to distraction and reduced STM capacity
seem to characterize the cognitive profile of anxious individuals. However,
permeability between these different functions leads to a need to examine their
independent effects as well as how they are related. The same idea has been explored
Hirsh, Clark and Mathews (2006) who propose the combined cognitive bias
hypothesis, suggesting that different cognitive biases associated with anxiety
disorders may mutually influence each other and also interact together to produce
multiplicative effects on levels of anxiety. The purpose of this study is thus to
examine independent and related effects of three different cognitive functions
(attentional bias, memory, and distractibility) within the same experiment. In
addition, we wanted to examine how these functions operate in the verbal and
visuospatial domains, in relation to anxiety and worry. An additional goal was to
examine whether different cognitive profiles characterize clinical levels of anxiety
relative to non-clinical samples. This study is part of a larger research project
examining cognitive correlates of GAD.
To examine these questions, we use an innovative method, including indexes of
cognitive functioning based on performance in the irrelevant speech paradigm and
the visuospatial sandwich paradigm. In the irrelevant speech paradigm, participants
must perform a difficult primary task, involving the serial recall of letters. On some
trials, irrelevant speech is presented and this has a deleterious effect on performance
compared to a control condition without irrelevant speech. This paradigm is well
296 M.-L.B. Lapointe et al.

established in cognitive experimental psychology (e.g., Alley & Greene, 2008; Bell,
Mund, & Buchner, 2011; Colle & Welsh, 1976) but has not been used with an anxious
population. By using neutral and threatening irrelevant speech, we can examine both
distractibility (irrelevant neutral vs. control) and attentional bias toward threat
(irrelevant threatening vs. irrelevant neutral). In addition, this paradigm provides a
direct measure of verbal STM capacity that can be derived by looking at recall
performance in the control condition.
The original sandwich paradigm (Hitch, 1975) was developed to evaluate the
interference effect in verbal STM. Results using this task have shown that the
inclusion of irrelevant stimuli ‘‘sandwiched’’ between to-be-remembered items
decreases recall performance (Baddeley, Papagno, & Andrade, 1993; Nicholls &
Jones, 2002). Tremblay, Nicholls, Parmentier and Jones (2005) proposed an
adaptation of the sandwich paradigm to evaluate visuospatial STM, in which seven
to-be-remembered black dots are presented in succession and in different locations
on a computer screen. On some trials, irrelevant dots are presented in the center of
the display, sandwiched between to-be-remembered items. This paradigm is used to
investigate both visuospatial STM and distractibility in spatial coding and has not
yet been studied in relation to anxiety. The use of this paradigm allows us to examine
distractibility when the need is to inhibit processing of neutral visual distracters.
We examine performance on these two tasks as a function of state anxiety, the
propensity to worry, and an index of clinical levels of anxiety. While attentional bias
has been documented as a function of both state and trait anxiety, and though both
measures are strongly correlated, there is some evidence that state anxiety may be the
strongest predictor of the two (Bar-Haim et al., 2007; Dresler, Me´riau, Heekeren, &
van der Meer, 2009). We also include a specific measure of worry because this may be
particularly detrimental for verbal information processing. Finally, we include a
questionnaire that is used to determine whether individuals meet GAD diagnostic
criteria. This measure is intended to examine whether cognitive effects associated
with anxiety and worry generally are also likely to characterize clinical anxiety.
Because comparing clinical and non-clinical levels of anxiety was one of the goals of
the study, we oriented recruitment to ensure our sample would include a sizeable
portion of participants likely show high levels of anxiety and potentially meet the
diagnostic criterion for GAD.
Method
Participants
The study was advertised on campus at Universite´Laval by posting ads in different
departments. There were general ads presenting the project as an investigation of
cognitive function (including distractibility, attention, and memory). Other ads
specifically invited highly anxious individuals to participate in a study examining
how anxiety and worry are related to cognitive function. The use of the latter was to
ensure that our sample would comprise a large enough subgroup of individuals with
high levels of anxiety, enabling us to examine whether a specific profile was likely to
characterized clinically anxious individuals.
Individuals under 18 or over 65 years old were excluded. Individuals who obtained
a score of 14 or over on the Beck depression inventory-II (BDI-II; Beck, Steer, &
Anxiety, Stress, & Coping 297

Brown, 1996) were also excluded because studies have demonstrated that people
suffering from depressive symptoms have different cognitive perturbations than
anxious individuals (Logan & Goetsch, 1993; Mogg, Bradley, & Williams, 1995).
Seventy-seven individuals participated in the study. The sample included 62
females and 15 males, aged between 19 and 31 years (M22.44, SD3.15).
Participants received a small honorarium for taking part in the study.
Materials and procedure
Questionnaires
The Statetrait anxiety inventory (STAI Form Y-1; Spielberger, 1983) is a measure
composed of 20 items evaluating current state anxiety level. Scores can range from 20
to 80; the average in our sample was 34.31 (SD 10.89) and actual scores ranged
from 20 to 62. The French version has very good internal consistency (Cronbach’s
alpha.90) and excellent construct validity similar to the English version (Gauthier
& Bouchard, 1993).
The Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec,
1990) includes 16 items which evaluate the tendency to worry in adults, rated on a
5-point Likert scale. The French version used in this study has excellent testretest
reliability (4 weeks; r.86) and very good internal consistency (Cronbach’s
alpha.92) (Gosselin, Dugas, Ladouceur, & Freeston, 2001). The average score in
our sample was 45.31 (SD16.75) and scores ranged from 0 to 74, out of a possible
maximum of 80.
The Worry and Anxiety Questionnaire (WAQ; Dugas et al., 2001) includes
questions evaluating GAD diagnostic criteria according to DSM-IV (APA, 1994),
using 9-point Likert scales (08). The WAQ has adequate specificity and sensitivity,
as well as sufficient testretest reliability (9 weeks; r.79; Dugas et al., 2001). The
WAQ is used to determine whether individuals are likely to meet GAD diagnostic
criteria, according to the profile of responses to the different items. We used this
method to identify participants considered likely to suffer from GAD according to
this measure. The dichotic classification (GAD/non-GAD) was used for discriminant
analyses to determine whether specific cognitive profiles characterized clinically
anxious individuals.
Irrelevant speech paradigm
The irrelevant speech paradigm is a measure of selective attention. It is used to
investigate the ability to inhibit irrelevant auditory speech presented during a verbal
serial recall task. This paradigm is a modified version of the task initially created by
Colle and Welsh (1976). Seven-item lists of to-be-remembered items were constructed
using random combinations of the consonants F, K, L, M, R, T, and V. The letters
were presented in uppercase 54-point Tahoma font and were black on a white
background in the center of a 17 17 cm matrix on a PC computer screen. Letters
were displayed at a rate of one letter per second with a one-second period between
each letter, followed by a 10 sec wait before recall. The word ‘‘recall’’ then appeared
on the screen, prompting participants to write the letters in the order of presentation.
After recall, participants used the mouse to click on ‘‘start’’ to generate a new trial.
298 M.-L.B. Lapointe et al.

In the irrelevant speech conditions, words were presented auditorily during the
presentation of the letters and the 10 sec rehearsal periods. All words had a duration
of one second and were composed of two phonemic syllables. They were spoken by a
male voice, digitally recorded with 12 bit resolution, and played back via headphones
at approximately 70 dB. Words had equal frequency of use in the French language
according to Baudot (1992).
To obtain neutral and threatening words, we asked people from the general
population and psychologists specialized in anxiety disorders to suggest stimuli with
high threatening or neutral valence. The most frequently suggested and appropriate
stimuli (according to Baudot) were then selected, yielding 34 threatening words and
34 neutral words (see Table 1). Two sequences of 17 words were recorded and
presented sequentially, starting with one or the other randomly.
There were two practice trials in the control condition prior to the experimental
trials. A total of 36 experimental trials were presented, 12 trials in each of the three
conditions: control, irrelevant neutral speech, and irrelevant threatening speech. The
dependent measure was the proportion of letters recalled in correct position. The
task lasted approximately 25 minutes.
Visuospatial sandwich paradigm
The visuospatial sandwich paradigm measures participant’s ability to process task-
relevant information and to ignore task-irrelevant information in the visuospatial
domain. The task consists in presenting a sequence of dots at different spatial
positions on a computer screen, and then requiring participants to reproduce the
sequence. On every trial, seven to-be-remembered black dots of .85 cm in diameter
were presented in succession in different locations within a 17 17 cm matrix (which
was not visible to participants at any point). The coordinates of the to-be-
remembered dots were randomly generated, with the restriction that the center of
successive dots had to be separated by at least 3 cm and no to-be-remembered dot
could appear closer than 3 cm to the center point of the presentation window.
In trials with distraction (sandwich condition), irrelevant dots (identical in size
and color to the to-be-remembered dots) always occupied the center point position
of the display. Participants were asked to fixate on the screen between the
presentation of to-be-remembered items but told that any dot presented in the
center of the screen was irrelevant to the memory task and should be ignored.
Following the presentation of the last to-be-remembered dot in each trial, all to-
be-remembered dots reappeared simultaneously in the same spatial locations where
they had originally been presented. Participants were asked to reconstruct the order
of the sequence using the mouse to click on the dots in the order in which they were
presented. To indicate that a response had been recorded, the color of a selected dot
changed from black to green. Mouse clicks on a dot already selected were not
processed. Written instructions encouraged participants to respond as quickly and as
accurately as possible and informed them that once an item had been selected it
would not be possible to alter the selection. The responses were scored automatically
with respect to serial position and condition. The dependent measure is the
proportion of items recalled in correct position.
There were two control conditions in the visuospatial sandwich task: a condition
in which seven black dots were displayed at a rate of one dot per second (slow
Anxiety, Stress, & Coping 299

condition: 700 ms ‘‘on,’’ 300 ms ‘‘off’’), and a condition in which the sequence of
dots was displayed at a rate of two dots per second (fast condition: 350 ms ‘‘on,’’ 150
ms ‘‘off’’). In the sandwich condition, timing of to-be-remembered items was the
same as in the slow condition but irrelevant items (on for 700 ms) were added 300 ms
after the offset of each to-be-remembered item. The sequence always ended with the
presentation of a to-be-remembered item. The faster condition was included to
Table 1. Threatening and neutral words employed in the
irrelevant speech paradigm.
Threatening words Neutral words
First sequence
Deluge (flood) Pronom (pronoun)
Carcasse (carcass) Divan (couch)
Poignard (dagger) Fossile (fossil)
Cachot (dungeon) Camper (camp)
Tonnerre (thunder) Clavier (keyboard)
Poison (poison) Veston (jacket)
Bagarre (fight) Canton (canton)
Torture (torture) Crayon (pencil)
Divorce (divorce) Panier (basket)
Panique (panic) Gratuit (free)
Voleur (thief) Bagage (luggage)
Re´volte (rebellion) Rural (rural)
E
´clair (lightning) Cure´(reverend)
E
´chec (failure) Cuisine (kitchen)
Tornade (Tornado) Bazar (junk)
Volcan (volcano) Baril (barrel)
Arme (weapon) Chiffre (digit)
Second sequence
Gangre`ne (gangrene) Alpin (alpine)
Vampire (vampire) Biscotte (biscuit)
Cercueil (coffin) Annexe (annex)
Violer (rape) Carotte (carrot)
Hante´(haunt) E
´vier (sink)
Morbide (morbid) Festin (feast)
Cauchemar (nightmare) Feˆter (celebrate)
Sanglant (bloody) Chimie (chemistry)
Fusil (gun) Statue (statue)
Fantoˆme (phantom) Cousin (cousin)
Bandit (bandit) Profil (profile)
Monstre (monster) Cristal (crystal)
Diable (devil) Carton (cardboard)
Meurtre (murder) Plateau (tray)
Folie (madness) Concert (concert)
Prison (jail) Tissu (tissue)
Esprit (ghost) Valeur (value)
Note: Words in parentheses are English translations of the words
presented in the original French version of the task.
300 M.-L.B. Lapointe et al.

control for the fact that the irrelevant dots in the sandwich condition may have made
the rate of display appear faster. The faster condition would allow us to ensure that
the distractibility effect was due to the irrelevant items rather than the rate of display.
There were three practice trials (one in each condition) followed by 45
experimental trials: 30 in the control condition (15 with the slower rate and 15
with the faster rate) and 15 in the sandwich condition. The order of presentation
was quasi-randomized with the restriction that no more than three trials of the
same condition could be presented successively. This task lasted approximately
25 minutes.
General procedure
Participants were tested individually. After signing the consent form, participants
completed the questionnaires, followed by the two experimental tasks: the irrelevant
speech paradigm and the visuospatial sandwich paradigm. The experimental tasks
were counterbalanced. For this, participants were seated approximately 50 cm away
from the computer screen. Finally, participants completed a manipulation check.
They were asked to rate the subjective threat value of each of the 68 words
(threatening and neutral) on a 5-point Likert scale (0not threatening, to
4extremely threatening).
Analyses
Three indices of cognitive function were calculated: attentional bias, distractibility,
and STM capacity. The attentional bias index provided an estimate of the influence
of threatening irrelevant stimuli on performance. This index corresponds to the
difference between recall performance in the neutral and threatening conditions in
the irrelevant speech paradigm, in proportion to the number of items recalled in
the neutral condition (used as a baseline). The distractibility index was used to
indicate the influence of irrelevant neutral information on recall performance.
A mean was calculated for: (1) the difference between performance in the silence
condition and the neutral distracter condition in the irrelevant speech paradigm
and (2) the difference between performance in the slower control condition and the
condition with distracters (sandwich condition) in the visuospatial sandwich
paradigm.
Again both these differences are calculated in proportion to the level of
performance in the condition without distracters. Finally, the memory index, which
corresponds to the memory span baseline, was calculated as the mean of the control
conditions of each paradigm, that is, the silence condition for the irrelevant speech
paradigm and the slower control condition for the visuospatial sandwich paradigm.
The formulae used for index calculation are presented in Table 2. We examined the
link between anxiety/worry and each of the three main cognitive functions,
controlling for the effect of the other cognitive functions, using hierarchical
regression analyses. We also compared participants likely to meet the diagnostic
criteria for GAD (based on the WAQ) to others to examine if there was a specific
cognitive profile associated with clinical levels of anxiety.
Anxiety, Stress, & Coping 301

Results
Manipulation check
At-test confirmed that threatening words obtained significantly higher threat ratings
(M1.76, SD.78) than neutral words, (M.11, SD .21), t(76)19.46, pB.001.
We also verified that participants classified as likely to meet the criteria for GAD
based on the WAQ, the clinically anxious group (N26, 34%), differed from the rest
of the sample (N51) on reported state anxiety, t(75)6.1, pB.01 (M43.1,
SD7.6, and M29.8, SD9.6, respectively) and level of worry t(75) 9.6, pB.01
(M52.2, SD6.8, and M21.6, SD15.5, respectively).
A repeated-measures ANOVA confirmed that performance in the visuospatial
sandwich paradigm differed depending on condition, F(2, 152)102.53, pB.01.
Bonferoni corrected pairwise comparisons indicated that performance in the
sandwich condition (M.47, SD.15) differed from both the slow control
condition (M.63, SD.14), t(76)13.59, pB.01 and the fast control condition,
(M.55, SD.14), t(76)7.61, pB.01, confirming the effect of additional
information was not simply due to the rate of presentation.
The same strategy was used to examine overall performance in the irrelevant
speech paradigm, confirming a main effect of condition, F(2, 152) 101.61, pB.01.
Average proportion of items recalled correctly was greater in the control condition
(M.73, SD.15) compared to conditions presenting irrelevant neutral speech
(M.59, SD.17), t(76)11.09, pB.01, and irrelevant threatening speech,
(M.58, SD.16), t(76)12.40, pB.01.
Correlations between cognitive functions, anxiety, and worry
Correlations are presented in Table 3. Results show a significant positive correlation
between the attentional bias index and scores of state anxiety and worry. Scores on
these latter two questionnaires also correlated (r.73, pB.01). The distractibility
index was not significantly correlated with any questionnaire. The global memory
index was significantly negatively correlated with both questionnaires. An elevated
tendency to worry and increased state anxiety were linked to poorer memory
performance. Specific analyses separating the two components of STM revealed that
Table 2. Formulae for the calculation of attentional bias, distractibility, and memory indexes in
the irrelevant speech and visuospatial sandwich paradigms.
Index Formula
Attentional
bias (only
for
irrelevant
speech
paradigm)
ðNeutral conditionThreatening threatening conditionÞ
Neutral condition
Distractibility
ðSilence conditionneutral distracter conditionÞ
Silent condition Irrelevant speech paradigm þðSlow control conditionSandwich sandwich conditionÞ
Slow control condition visuospatial sandwich paradigm
2
Memory Slow condition visuospatial sandwich paradigm þSilence condition irrelevant speech paradigm
2
302 M.-L.B. Lapointe et al.

Table 3. Correlations between indexes, perception of threat value and questionnaires.
Indexes
Questionnaires
Distractibility
index
Attentional bias
index
Memory
index
Verbal
memory
Visuospatial
memory
Rating of threatening
words
a
Rating of neutral
words
State anxiety
(STAI)
.01 .33** .38** .42** .16 .46** .28*
Worry (PSWQ) .16 .25* .29** .17 .29* .49** .16
N77 participants.
STAI, State and Trait Anxiety Inventory (Form Y-1); PSWQ, Penn State Worry Questionnaire.
a
Rating correspond to the rating of the threat value of threatening and neutral words used in the irrelevant speech paradigm.
*pB.05; **pB.01.
Anxiety, Stress, & Coping 303

verbal STM was negatively related to state anxiety while visuospatial STM was
negatively related to worry. Finally, evaluation of threatening words was significantly
correlated with state anxiety and worry, as well as attentional bias (r.26, pB.05). In
addition, threat ratings of neutral words were also linked with state anxiety.
Regression analyses
To investigate the unique predictive ability of individual variables for different
cognitive function indices we conducted hierarchical regression analyses. Table 4
presents the unique contribution (sR
2
) of all steps of the regression. When a step
includes more than one variable, the Beta indicates the strength of this variable in
that regression step. Our strategy was first to examine the extent to which each
cognitive function was predicted by the other cognitive functions, to assess
permeability. When relevant, we also included subjective evaluation of the stimuli.
We then included the anxiety-related measures in one order (model 1: worry first and
state anxiety second) and the other (model 2: state anxiety first, then worry) to
examine the unique contribution of each, above and beyond what could be predicted
by cognitive factors and by the other anxiety-related variable. Considering the high
correlation between worry and state anxiety, there was a potential problem with
multi-collinearity in the regression analyses. This was prevented by using scores
regressed to mean for all variables, as proposed by Aiken and West (1991).
For the global memory index, results show that attentional bias and distract-
ibility were not significant predictors. State anxiety appears to be most important as
it accounted for a significant amount of variance, whether entered before or after
worry. Thus, results show that overall memory is relatively independent from the
other two cognitive variables assessed, and more strongly linked with state anxiety
than worry.
Similar analyses were conducted to examine predictors of distractibility, this time
adding participants’evaluation of neutral stimuli in the first step. This was included
because the distractibility index is calculated from performance on neutral stimuli
and because of the possibility that neutral stimuli may not be equally neutral for all
participants. Results showed that of the two cognitive indices, attentional bias
predicted a significant part of the variance in distractibility and that this relationship
was negative. In this case, worry was a more important predictor than state anxiety.
Worry predicted a significant amount of additional variance in distractibility when
entered in a second or third step, while this was not true of state anxiety. These
results suggest that worry, but not state anxiety, explains some unique variance in
distractibility above and beyond what is accounted for by cognitive indices.
Attentional bias toward threat was examined using the same strategy, this time
entering evaluation of threatening stimuli in the first step. Distractibility accounted
for some of the variance in attentional bias, the relationship being again negative.
When either worry or state anxiety was added in a second step, this resulted in a
significant increase in variance accounted for. The individual predictors did not,
however, reach significance. When entered in a third step, worry and state anxiety did
not explain additional unique variance, suggesting that the variance explained was
common to both variables.
To summarize, while permeability between cognitive functions (especially
distractibility and attentional bias) explained some of the variance in each index,
304 M.-L.B. Lapointe et al.

Table 4. Hierarchical regression models.
Dependent variable Step Predictors (bin final model) Fchange sR
2
Memory model 1 1 Attentional bias (.18) F(2, 74) 0.2 .00
Distractibility (.05)
2 Worry (.06) F(1, 73) 8.1** .10
3 State anxiety (.40)* F(1, 72) 5.9* .17
Memory model 2 1 Attentional bias (.18) F(2, 74) 0.2 .00
Distractibility (.05)
2 State anxiety (.40)* F(1,73) 14.6** .17
3 Worry (.06) F(1, 72) 0.11 .17
Distractibility model 1 1 Evaluation of neutral words (.08) F(3, 73) 3.7* .13
Memory (.05)
Attentional bias (.40)**
2 Worry (.35)* F(1, 72) 6.8* .20
3 State anxiety (.08) F(1, 71) 0.2 .20
Distractibility model 2 1 Evaluation of neutral words (.08) F(1, 73) 3.7* .13
Memory (.05)
Attentional bias (.40)**
2 State anxiety (.08) F(4, 72) 1.9 .15
3 Worry (.35)* F(5, 71) 5.0* .20
Attentional bias model 1 1 Evaluation of threatening words (.03) F(1, 73) 5.7* .19
Memory (.15)
Distractibility (.38)**
2 Worry (.13) F(4, 72) 5.2* .25
3 State anxiety (.26) F(5, 71) 2.6 .27
Anxiety, Stress, & Coping 305

Table 4 (Continued )
Dependent variable Step Predictors (bin final model) Fchange sR
2
Attentional bias model 2 1 Evaluation of threatening words (.08) F(2, 73) 5.7** .19
Memory (.15)
Distractibility (.38)**
2 State anxiety (.26) F(4, 72) 7.3** .27
3 Worry (.13) F(5, 71) 0.6 .27
*pB.05; ** pB.01
306 M.-L.B. Lapointe et al.

unique additional predictive value was observed for anxiety-related variables.
Specifically, memory was particularly related to state anxiety, while distractibility
was mostly related to worry. State anxiety and worry together were linked with
attentional bias, which interestingly was not directly linked with subjective
evaluations of threatening stimuli.
Discriminant analysis
Discriminant analyses were performed in order to determine whether indices of
cognitive functioning and subjective threat evaluations could be used to classify
participants as clinical or non-clinical on the basis of participants’scores on the
WAQ. With this questionnaire, participants can be defined as fulfilling clinical
criteria (GAD) or not fulfilling clinical criteria (non-GAD) according to DSM-IV
(APA, 1994).
A first discriminant function was conducted on the sample to identify if it was
possible to discriminate between groups using all variables: the three cognitive
indices (attentional bias, distractibility, working memory) and the threat evaluation
index. All variables were entered in the same block. The discriminant function
showed a significant global Wilks’lambda, L.75, x
2
(4, N77)21.20, pB.01,
suggesting that these variables can be used to distinguish between the two groups as
defined by WAQ scores. When the four different variables were considered separately
in distinct discriminant analyses, results indicated that distractibility (p.15) and
memory indices (p.22) were not significant. Results for the attentional bias index,
(L.92, x
2
(1, N77)6.02, pB.01), and the threat evaluation index, (L.89,
x
2
(1, N77)8.76, pB.01), were both significant.
Discussion
This study showed that state anxiety is related to impairments in STM while worry is
more closely related to distractibility. Attentional bias is linked with the common
variance shared between state anxiety and worry and particularly characteristic of
clinical levels of anxiety. Globally these findings suggest that though there is some
overlap between the cognitive correlates of anxiety, state anxiety, and worry do also
contribute to explain additional variance that is not accounted for by the common
effects of these cognitive functions.
The attentional bias toward threat in anxiety is well documented in the literature
(see Bar-Haim et al., 2007). Several theoretical models of clinical anxiety (Mathews
& Mackintosh, 1998; Mogg & Bradley, 1998; Williams, Watts, MacLeod, & Mathews,
1988) or fear processing (O
¨hman & Mineka, 2001) include attentional bias toward
threat as a central feature. Our findings offer two important contributions with
respect to this attentional bias. A first novel contribution is that our results clearly
establish that while anxious individuals show an attentional bias toward threat and
also show increased threat evaluations, the former does not result from the latter.
Anxious individuals evaluate stimuli as more threatening than non-anxious
individuals. Given this, the interpretation of the attentional bias toward threat is
ambiguous. Does this reflect a genuine allocation of resources toward threat or does
it reflect the fact that even neutral stimuli are perceived as more threatening by
anxious individuals? Our analyses show that when controlling for the differential
Anxiety, Stress, & Coping 307

threat evaluations, anxiety still contributes to explain a significant portion of
additional variance in attentional bias toward threat, suggesting there is a genuine
bias in the allocation of information processing resources, rather than an artifact of
differential threat perception.
The second contribution of our results is that we are able to rule out the
possibility that the attentional bias toward threat results purely from increased
distractibility. There was a link between distractibility and attentional bias, but this
link was negative. This was an unexpected finding as we had anticipated that
increased distractibility may explain some of the attentional bias toward threat.
Instead, greater levels of distractibility were related to decreased attentional bias
toward threat. This may be because the attentional bias is a relative measure,
essentially comparing the interference from threatening and neutral stimuli.
Individuals who are highly distractible may be equally swayed by both types of
stimuli, resulting in a kind of ceiling effect. Less distractible individuals may show
increased interference only from threatening stimuli that are more powerful at
capturing attention. Despite this unexpected finding, an important finding was that
state anxiety and worry provided additional predictive value even when the effect of
distractibility was accounted for. This suggests that the robust attentional bias
toward threat linked with anxiety is circumscribed to threat and inherently linked
with the nature of the stimuli, rather than being a broad ranging effect.
Interestingly, our results show that distractibility is more uniquely linked with
worry than with state anxiety. Though these two constructs (worry and state anxiety)
are highly correlated, our analyses nevertheless suggest specific effects of each. It
appears that STM decrements are more uniquely tied to state anxiety while
distractibility is more closely linked with worry. Concerning distractibility, our
results also rule out the possibility that overall effects of distractibility are due to a
confound in the nature of the stimuli. If the neutral stimuli are perceived as more
threatening by anxious or high worriers, coupled with an attentional bias, this could
account for the observed increased in distractibility. Results show that even when
perception of neutral stimuli and attentional bias were accounted for, worry still
explained additional variance in distractibility.
The correlations do show that the evaluation of neutral stimuli was linked with
state anxiety. More anxious individuals tended to evaluate not only threatening items
but also neutral items as more threatening than less anxious individuals. This
highlights how important it is to have an assessment of the threat value of stimuli
when attempting to characterize generic cognitive function. Conclusions about the
impact of anxiety on processing of neutral information must include careful
manipulation checks to ensure that stimuli being presented are indeed subjectively
neutral, even to high-anxious individuals.
Discriminant analyses demonstrated that amongst the cognitive alterations
examined the attentional bias index and subjective threat evaluations can be used to
distinguish between clinical and non-clinical groups. This confirms that allocation of
cognitive resources toward threat in information processing and alterations in the
perceived threatening value of stimuli are robust central features of anxiety disorders. It
also suggests that the features that are associated with non-pathological anxiety are
also characteristic of clinical levels. We must, however, interpret these findings
concerning the comparison between the two groups of participants carefully. The
participants likely to meet the criteria forGAD were self-selected and proactive, as they
308 M.-L.B. Lapointe et al.

responded to advertisement specifically targeting individuals with anxiety problems
and thus may not be representative. Importantly, though the groups differed,
participants in the group likely to meet diagnostic criteria based on the WAQ showed
a group average state anxiety score of 43 on a scale with a maximum value of 80. This is
modest and suggests that the group presented relatively mild levels of clinically
significant anxiety.
Our findings have implications for current models of anxiety and cognitive
function. One prominent example is the Attentional Control Theory (Eysenck,
Derakshan, Santos, & Calvo, 2007). Broadly, this theory suggests that anxiety leads to
lower cognitive efficiency because it is associated with impairments in executive
function, particularly the ability to inhibit distracters and shifting. Anxious individuals
are less good at maintaining goal-focused attention and are more easily distracted by
irrelevant stimuli (internal or external), especially threatening stimuli. Processing
distracters take up processing capacity. Our results are consistent with many aspects of
this model. Our results show important effects of worry on distractibility, and confirm
that threatening stimuli are more difficult to ignore for anxious individuals. Results
also show STM storage deficits related to anxiety. All these findings are consistent with
the ACT theory. One important additional issue is that the ACT model does not really
differentiate between anxiety and worry. There is an assumption that worry is the
crucial process mediating the effect of anxiety on cognitive function. Our results
suggest that future versions of the model may gain from making finer distinctions
between components of anxiety (for instance teasing out state anxiety and worry) as
separate components may be differentially related to specific cognitive functions.
Another aspect we investigated is whether there are specific patterns of effects for
verbal and visuospatial modalities. This is important from a cognitive point of view
given it is well established that these represent specific functions relying on
independent mechanisms (Jonides et al., 1996). We observed that state anxiety
correlated more strongly with verbal STM while worry was more strongly related to
visuospatial STM. These counterintuitive results seem to contradict the view of
Eysenck (1979), who suggests that worries mainly overload verbal STM because of
their verbal nature. Results also clash with some empirical findings that showed strong
links between worry and STM in the verbal, but not visual domain (Leigh & Hirsch,
2011). Our finding must be interpreted with caution as the WAQs were themselves
highly correlated (r.70). Furthermore, the visual and verbal memory tasks presented
different characteristics that may differentially relate to anxiety. Ideally, verbal and
visual tasks with the same structure should be used to make direct comparisons.
Nevertheless our findings are consistent with some recent empirical and theoretical
work on imagery and worry. Borkovec’s Cognitive Avoidance Theory (Borkovec,
Alcaine, & Behar, 2004) proposes that high-anxious individuals engage in verbal
thought (worry) to avoid negative images. Hirsch, Hayes, Mathews, Perman and
Borkovec (2012) have recently examined the occurrence of imagery and verbal
thoughts during periods of worry and other periods. They found that individuals
suffering from GAD exhibit truncated imagery; the mental images they experience are
generally briefer, during worry and other periods. The researchers interpreted this
finding as evidence that individuals with GAD adopt more verbal thinking styles
generally. One possibility is that this increased reliance on verbal processing results
from poorer visuospatial processing, an interpretation that our results concerning the
link between worry and visuospatial STM would be consistent with. Thus, though the
Anxiety, Stress, & Coping 309

link between worry and visuospatial processing may not be intuitive, it is consistent
with some findings in the literature and will merit additional future consideration.
Our study presents certain limitations. We did not include threatening distracters
in the visuospatial task, which prevents us from generalizing our findings concerning
attentional bias toward threat to the visuospatial modality. We included only a
measure of state anxiety, as previous research had shown that this is the more
important predictor of attentional bias, compared to trait anxiety. However,
including a measure of trait anxiety in future studies will allow researchers to
examine whether state and trait anxiety are differentially related to distractibility and
STM capacity. Finally, future studies with more numerous clinically anxious
participants for whom a valid diagnostic will be obtained will be necessary to
provide stronger conclusions on the features that characterize clinical levels of
anxiety.
Despite these limitations, this study has provided important contributions,
especially by examining three major cognitive dysfunctions associated with anxiety
simultaneously. Attentional bias toward threat, distractibility, and impairments in
STM capacity are the three cognitive dysfunctions most studied in the anxiety
literature. Our findings confirm the important links that exist between anxiety and
cognitive functioning.
References
Aiken, L.S., & West, S.G. (1991). Multiples regressions: Testing and interpreting interactions.
Thousand Oaks, CA: Sage Publication.
Alley, T., & Greene, M.E. (2008). The relative and perceived impact of irrelevant speech, vocal
music and non-vocal music on working memory. Current Psychology,27(4), 277289.
American Psychiatric Association [APA]. (1994). Diagnostic and statistical manual of mental
disorders (4th ed.). Washington, DC: Auteur.
Baddeley, A., Papagno, C., & Andrade, J. (1993). The sandwich effect: The role of attentional
factors in serial recall. Journal of Experimental Psychology: Learning, Memory, and
Cognition,19(4), 862870.
Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M.J., & van Ijzendoorn,
M.H. (2007). Threat-related attentional bias in anxious and nonanxious individuals: A
meta-analytic study. Psychological Bulletin,133(1), 124. doi:10.1037/0033-2909.133.1.1.
Baudot, J. (1992). Fre´quences d’utilisation des mots en franc
¸ais e´crit contemporain. Montre´al:
Presses de l’Universite´de Montre´al.
Beck, A.T., Steer, R.A., & Brown, G.K. (1996). Beck depression inventory manual (2nd ed). San
Antonio, TX: Psychological Corporation.
Bell, R., Mund, I., & Buchner, A. (2011). Disruption of short-term memory by distractor
speech: Does content matter? The Quarterly Journal of Experimental Psychology,64(1),
146168.
Blanchette, I. (2006). Snakes, spiders, guns, and syringes: How specific are evolutionary
constraints on the detection of threatening stimuli? The Quarterly Journal of Experimental
Psychology,59(8), 14841504. doi:10.1080/02724980543000204.
Blanchette, I. & Richards, A. (2010). The influence of affect on higher level cognition: A
review of research on interpretation, judgement, decision making and reasoning. Cognition
and Emotion,24(4), 561595. doi:10.1080/02699930903132496.
Borkovec, T.D., Alcaine, O., & Behar, E. (2004). Avoidance theory of worry and generalized
anxiety disorder. In R.G. Heimberg, C.L. Turk, & D.S. Mennin (Eds.), Generalized anxiety
disorder: Advances in research and practice (pp. 77108). New York: The Guilford Press.
Borkovec, T.D., Robinson, E., Pruzinsky, T., & DePree, J.A. (1983). Preliminary exploration of
worry: Some characteristics and processes. Behaviour Research and Therapy,21,916.
310 M.-L.B. Lapointe et al.

Bradley, B.P., Mogg, K., Falla, S.J., & Hamilton, L.R. (1998). Attentional bias for threatening
facial expressions in anxiety: Manipulation of stimulus duration. Cognition and Emotion,
12(6), 737753. doi:10.1080/026999398379411.
Bradley, B.P., Mogg, K., & Williams, R. (1994). Implicit and explicit memory for emotional
information in non-clinical subjects. Behaviour Research and Therapy,32(1), 6578.
doi:10.1016/0005-7967(94)90085-x.
Colle, H.A., & Welsh, A. (1976). Acoustic masking in primary memory. Journal of Verbal
Learning & Verbal Behavior,15(1), 1731. doi:10.1016/s0022-5371(76)90003-7.
Corsi, P.M. (1973). Human memory and the medial temporal region of the brain. Dissertation
Abstracts International,34(2-B), 891.
Crowe, S.F., Matthews, C., & Walkenhorst, E. (2007). Relationship between worry, anxiety
and thought suppression and the components of working memory in a non-clinical sample.
Australian Psychologist,42(3), 170177. doi:10.1080/00050060601089462.
Darke, S. (1988). Anxiety and working memory capacity. Cognition and Emotion,2(2),
145154. doi:10.1080/02699938808408071.
Derakshan, N., & Eysenck, M.W. (1998). Working memory capacity in high trait-anxious and
repressor groups. Cognition and Emotion,12(5), 697713. doi:10.1080/026999398379501.
Dresler, T., Me´riau, K., Heekeren, H.R., & van der Meer, E. (2009). Emotional stroop task:
Effect of word arousal and subject anxiety on emotional interference. Psychological
Research/Psychologische Forschung,73(3), 364371.
Dugas, M.J., Freeston, M.H., Provencher, M.D., Lachance, S., Ladouceur, R., & Gosselin, P.
(2001). Le Questionnaire sur l’Inquie´tude et l’Anxie´te´. Validation dans des e´chantillons non
cliniques et cliniques [The worry and anxiety questionnaire: Validation in non-clinical and
clinical samples]. Journal de The´rapie Comportementale et Cognitive,11(1), 3136.
Eysenck, M.W. (1979). Anxiety, learning, and memory: A reconceptualization. Journal of
Research in Personality,13(4), 363385. doi:10.1016/0092-6566(79)90001-1.
Eysenck, M.W., & Byrne, A. (1992). Anxiety and susceptibility to distraction. Personality and
Individual Differences,13(7), 793798. doi:10.1016/0191-8869(92)90052-q.
Eysenck, M.W., Derakshan, N., Santos, R., & Calvo, M.G. (2007). Anxiety and cognitive
performance: Attentional control theory. Emotion,7(2), 336 353. doi:10.1037/1528-
3542.7.2.336.
Eysenck, M.W., & Graydon, J. (1989). Susceptibility to distraction as a function of personality.
Personality and Individual Differences,10(6), 681687. doi:10.1016/0191-8869(89)90227-4.
Eysenck, M.W., Payne, S., & Derakshan, N. (2005). Trait anxiety, visuospatial processing, and
working memory. Cognition and Emotion,19, 12141228.
Firetto, A.C., & Davey, H. (1971). Subjectively reported anxiety as a discriminator of digit
span performance. Psychological Reports,29, 98.
Fox, E. (1993). Attentional bias in anxiety: Selective or not? Behaviour Research and Therapy,
31(5), 487493. doi:10.1016/0005-7967(93)90129-i.
Fox, E. (1996). Selective processing of threatening words in anxiety: The role of awareness.
Cognition and Emotion,10(5), 449480. doi:10.1080/026999396380114.
Fox, E., Griggs, L., & Mouchlianitis, E. (2007). The detection of fear-relevant stimuli: Are
guns noticed as quickly as snakes? Emotion,7(4), 691696. doi:10.1037/1528-3542.7.4.691.
Fox, E., Russo, R., Bowles, R., & Dutton, K. (2001). Do threatening stimuli draw or hold
visual attention in subclinical anxiety? Journal of Experimental Psychology: General,130(4),
681700.
Gauthier, J., & Bouchard, S. (1993). Adaptation canadienne-franc
¸aise de la forme re´vise´edu
StateTrait Anxiety. Canadian Journal of Behavioural Science/Revue canadienne des sciences
du comportement,25(4), 559578.
Georgiou, G.A., Bleakley, C., Hayward, J., Russo, R., Dutton, K., Eltiti, S., & Fox, E. (2005).
Focusing on fear: Attentional disengagement from emotional faces. Visual Cognition,12(1),
145158. doi:10.1080/13506280444000076.
Gosselin, P., Dugas, M.J., Ladouceur, R., & Freeston, M.H. (2001). valuation des inquie´tudes:
Validation d’une traduction franc
¸aise du Penn State Worry Questionnaire. L’Ence´phale:
Revue de psychiatrie clinique biologique et the´rapeutique,27(5), 475484.
Hayes, S., Hirsch, C., & Mathews, A. (2008). Restriction of working memory capacity during
worry. Journal of Abnormal Psychology,117(3), 712717. doi:10.1037/a0012908.
Anxiety, Stress, & Coping 311

Hirsch, C.R., Clark, D.M., & Mathews, A. (2006). Imagery and interpretations in social
phobia: Support for the combined cognitive biases hypothesis. Behavior Therapy,37, 223
236.
Hirsch, C.R., Hayes, S., Mathews, A., Perman, G., & Borkovec, T. (2012). The extent and
nature of imagery during worry and positive thinking in generalized anxiety disorder.
Journal of Abnormal Psychology,121(1), 238243.
Hitch, G.J. (1975). The role of attention in visual and auditory suffix effects. Memory &
Cognition,3(5), 501505.
Hodges, W.F., & Spielberger, C.D. (1969). Digit span: An indicant of trait or state anxiety?
Journal of Consulting and Clinical Psychology,33(4), 430434. doi:10.1037/h0027813.
Jonides, J., Reuter-Lorenz, P.A., Smith, E.E., Awh, E., Barnes, L.L., Drain, M., ...
Schumacher, E.H. (1996). Verbal and spatial working memory in humans. In D.L. Medin
(Ed.), The psychology of learning and motivation: Advances in research and theory Vol. 35
(pp. 4388). San Diego, CA: Academic Press.
Leigh, E., & Hirsch, C.R. (2011). Worry in imagery and verbal form: Effect on residual
working memory capacity. Behaviour Research and Therapy.49(2), 99105.
Logan, A.C., & Goetsch, V.L. (1993). Attention to external threat cues in anxiety states.
Clinical Psychology Review,13(6), 541 559. doi:10.1016/0272-7358(93)90045-n.
MacLeod, C., & Mathews, A. (1988). Anxiety and the allocation of attention to threat.
Quarterly Journal of Experimental Psychology,40, 653670.
MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emotional disorders. Journal
of Abnormal Psychology,95(1), 1520. doi:10.1037/0021-843x.95.1.15.
Mathews, A., & MacLeod, C. (2002). Induced processing biases have causal effects on anxiety.
Cognition and Emotion,16(3), 331354. doi:10.1080/02699930143000518.
Matthews, A., & Mackintosh, B. (1998). A cognitive model of selective processing in anxiety.
Cognitive Therapy and Research,22(6), 539560. doi:10.1023/a:1018738019346.
Mathews, A., & MacLeod, C. (2005). Cognitive vulnerability to emotional disorders.
Annual Review of Clinical Psychology,1(1), 167195. doi:10.1146/annurev.clinpsy.1.10
2803.143916.
Mathews, A., May, J., Mogg, K., & Eysenck, M. (1990). Attentional bias in anxiety: Selective
search or defective filtering? Journal of Abnormal Psychology,99(2), 166173. doi:10.1037/
0021-843x.99.2.166.
Meyer, T.J., Miller, M.L., Metzger, R.L., & Borkovec, T.D. (1990). Development and
validation of the Penn State Worry Questionnaire. Behaviour Research and Therapy,
28(6), 487495. doi:10.1016/0005-7967(90)90135-6.
Mialet, J.P. (2000). Les explorations cognitives de l’anxie´te´. In J.M.D.A.A.M. Van der Linden
(Ed.), La psychopathologie: Une approche cognitive et neuropsychologique (pp. 133160).
Marseille, FR: Solal e´diteurs.
Mogg, K., & Bradley, B.P. (1998). A cognitive-motivational analysis of anxiety. Behaviour
Research and Therapy,36(9), 809848. doi:10.1016/s0005-7967(98)00063-1.
Mogg, K., Bradley, B.P., Dixon, C., Fisher, S., Twelftree, H., & McWilliams, A. (2000). Trait
anxiety, defensiveness and selective processing of threat: An investigation using two
measures of attentional bias. Personality and Individual Differences,28, 10631107.
Mogg, K., Bradley, B.P., & Williams, R. (1995). Attentional bias in anxiety and depression:
The role of awareness. British Journal of Clinical Psychology,34(1), 1736.
Mogg, K., Mathews, A., Bird, C., & Macgregor-Morris, R. (1990). Effects of stress and
anxiety on the processing of threat stimuli. Journal of Personality and Social Psychology,
59(6), 12301237. doi:10.1037/0022-3514.59.6.1230.
Nicholls, A.P., & Jones, D.M. (2002). Capturing the suffix: Cognitive streaming in immediate
serial recall. Journal of Experimental Psychology: Learning, Memory, & Cognition,28,
1228.
O
¨hman, A., & Mineka, S. (2001). Fears, phobias, and preparedness: Toward an evolved
module of fear and fear learning. Psychological Review,108(3), 483522. doi: 10.1037/0033-
295x.108.3.483.
Pallack, M.S., Pittman, T.S., Heller, J.F., & Munson, P. (1975). The effect of arousal on Stroop
color-word task performance. Bulletin of the Psychonomic Society,6, 248250.
312 M.-L.B. Lapointe et al.

Reidy, J., & Richards, A. (1997). Anxiety and memory: A recall bias for threatening words in
high anxiety. Behaviour Research and Therapy,35(6), 531542. doi:10.1016/s0005-
7967(97)00001-6.
Richards, A., & French, C.C. (1990). Central versus peripheral presentation of stimuli in an
emotional Stroop task. Anxiety Research,3(1), 4149.
Richards, A., French, C.C., Johnson, W., & Naparstek, J. (1992). Effects of mood
manipulation and anxiety on performance of an emotional Stroop task. British Journal of
Psychology,83(4), 479491. doi:10.1111/j.2044-8295.1992.tb02454.x.
Shackman, A.J., Sarinopoulos, I., Maxwell, J.S., Pizzagalli, D.A., Lavric, A., & Davidson, R.J.
(2006). Anxiety selectively disrupts visuospatial working memory. Emotion,6(1), 4061.
doi:10.1037/1528-3542.6.1.40.
Sorg, B.A., & Whitney, P. (1992). The effect of trait anxiety and situational stress on working
memory capacity. Journal of Research in Personality,26(3), 235241. doi:10.1016/0092-
6566(92)90041-2.
Spielberger, C.D. (1983). Manual for the state-trait anxiety inventory (form Y). Palo Alto, CA:
Consulting Psychologist Press.
Tremblay, S., Nicholls, A.P., Parmentier, F.B.R., & Jones, D.M. (2005). Visual distraction and
visuo-spatial memory: A sandwich effect. Memory,13(34), 357363. doi:10.1080/
09658210344000422.
Van den Hout, M., Tenney, N., Huygens, K., & Merckelbach, H. (1995). Responding to
subliminal threat cues is related to trait anxiety and emotional vulnerability: A successful
replication of Macleod and Hagan (1992). Behaviour Research and Therapy,33(4), 451454.
doi:10.1016/0005-7967(94)00062-o.
Williams, J.M.G., Mathews, A., & MacLeod, C. (1996). The emotional stroop task and
psychopathology. Psychological Bulletin,120(1), 324. doi:10.1037/0033-2909.120.1.3.
Williams, J.M.G., Watts, F.N., MacLeod, C., & Mathews, A. (1988). Cognitive psychology and
emotional disorders. Oxford: John Wiley & Sons.
Wittchen, H.-U., & Hoyer, J. (2001). Generalized anxiety disorder: Nature and course. Journal
of Clinical Psychiatry,62(Suppl. 11), 15 19.
Yiend, J., & Mathews, A. (2001). Anxiety and attention to threatening pictures. The Quarterly
Journal of Experimental Psychology A: Human Experimental Psychology,54A(3), 665681.
doi:10.1080/02724980042000462.
Anxiety, Stress, & Coping 313

Copyright of Anxiety, Stress & Coping is the property of Routledge and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.