Journal of Anxiety Disorders, Vol. 12, No. 6, pp.

579–598, 1998
Pergamon Copyright  1998 Elsevier Science Ltd
Printed in the USA. All rights reserved
0887-6185/98 $–see front matter

PII S0887-6185(98)00035-8

REVIEW ARTICLE

Examining the Evidence for Anxiety-

Related Cognitive Biases in
Obsessive-Compulsive Disorder

Laura J. Summerfeldt, ph.d.

McMaster University, Hamilton, Ontario, Canada, and York University,
North York, Ontario, Canada

Norman S. Endler, ph.d., f.r.s.c

York University, North York, Ontario, Canada

Abstract—In light of current concerns about the diagnostic classification of obsessive-

compulsive disorder (OCD), this article critically examines recent experimental re-
search on anxiety-related cognitive biases in OCD in order to determine whether it pro-
vides grounds for OCD’s differentiation from other anxiety disorders. This small body
of work is found to be fraught with defects, anomalies, and inconsistencies. These find-
ings contrast dramatically with the robust results obtained with other clinical anxiety
disorders. When biases are in evidence it tends to be with a select group of subjects, that
is, those with contamination concerns. It is suggested that only this subtype of OCD, or
some core characteristic underlying it, may be associated with cognitive tendencies
comparable to those found in other anxiety disorders (i.e., biases at the attentional level
associated with the emotional tone, or content, of information). Cognitive tendencies in
other subtypes likely require different explanatory frameworks. This review provides
evidence for the partial uniqueness of OCD from other anxiety disorders.  1998 Else-
vier Science Ltd

The authors acknowledge the general support provided by the Social Sciences and Humanities Re-
search Council of Canada (SSHRC), in the form of a Doctoral Fellowship to the first author, and a
research grant (no. 3410-94-1473) to the second author.
Requests for reprints should be sent to Laura J. Summerfeldt, Ph.D., Anxiety Disorders Pro-
gram, Psychology Department, St. Joseph’s Hospital, 50 Charlton Ave. East, Hamilton, Ontario
L8N 4A6 Canada. E-mail: summerf@stjosham.on.ca
579
580 L. J. SUMMERFELDT AND N. S. ENDLER

Obsessive-compulsive disorder (OCD) has been designated as an anxiety-

based disorder in North America since the early days of the Diagnostic and
Statistical Manual of Mental Disorders (DSM; American Psychiatric Associa-
tion, 1952) nosological system, unlike many of the syndromes with which it
currently shares that diagnostic category. Despite this, recent years have wit-
nessed a growing number of allusions in the literature to problems with
OCD’s diagnostic classification (e.g., Clark, Watson, & Reynolds, 1995; En-
right, 1996; Enright & Beech, 1990, 1993a, 1993b; Hollander, 1993a, 1993c;
Reed, 1991; Stein & Hollander, 1993; Tynes, White, & Steketee, 1990).
Such nosological concerns warrant critical reappraisal of the empirical evi-
dence for commonalities, or incongruities, between OCD and other anxiety
disorders. The uniqueness of OCD is suggested by diverse sources, among
them reports of biochemical factors (Dolberg, Iancu, Sasson, & Zohar, 1996),
neuropsychological correlates (AlarHon, Libb, & Boll, 1994), symptom sever-
ity profiles (Cameron, Thyer, Nesse, & Curtis, 1986), Axis I comorbidity pat-
terns (e.g., Crino & Andrews, 1996), Axis II comorbidity patterns and preva-
lence (Mavisakkalian, Hamann, Haidar, & deGroot, 1993; Stein, Hollander,
Skodol, 1993), dimensional personality concomitants (Summerfeldt, Huta, &
Swinson, 1998) and a growing body of data pointing to its potential genetic,
neurochemical, and phenomenological links with such nosologically distinct
syndromes as tic-related disorders and trichotillomania—the hypothesized
obsessive compulsive spectrum disorders (see Hollander, 1993b; Oldham, Hol-
lander, & Skodol, 1996; Pauls, 1992; Rasmussen, 1994; Stein & Hollander,
1993).
This article critically examines another body of work—experimental re-
search on anxiety-related cognitive biases in OCD—in order to determine
whether it too provides grounds for the differentiation of OCD from other
anxiety disorders. This is a fruitful area in which to seek evidence. In the cog-
nitive experimental literature, the association of anxiety with a selective atten-
tional bias, favoring the processing of threatening information, has proven to
be robust and highly replicable. Its repeated demonstration in both clinical
anxiety disorders and nonclinical samples high in trait anxiety has led to it be-
ing regarded as a “well established” (Mogg, Mathews, Eysenck, & May, 1991,
p. 459) marker of anxiety states. As such, its unambiguous demonstration in
OCD would be pertinent to the growing debate about this disorder’s phenom-
enological and, potentially, etiological uniqueness from other anxiety syn-
dromes.
Cognitive findings should be considered a particularly valuable source of
information in the case of OCD. Despite the predominant emphasis on af-
fective factors implicit in its nosological placement, the cognitive features of
OCD have received more attention from researchers than those of any other
in the anxiety disorders category. The pervasiveness of this orientation was re-
cently evidenced by the unprecedented formation of an international collec-
tive of researchers with the mandate of collating, assessing, and developing
 COGNITIVE BIASES IN OCD 581

cognitively oriented self-report measures for use with this population (Obses-
sive Compulsive Cognitions Working Group, 1997). The characterization of
OCD in terms of an habitual style of managing information has historically
been a feature of clinical accounts (see Berrios, 1989; Reed, 1985, for reviews).
Unique cognitive tendencies have been described in many cases not only as
concomitants to the disorder, but as its principal feature (e.g., Janet, 1903;
Westphal, 1877). Contemporary clinical models of this disorder assign a piv-
otal role to idiosyncracies in cognition (e.g., Rachman, 1997; Salkovskis, 1985,
1989; see also Frost & Hartl, 1996). It is not surprising, then, that interest in in-
formation processing in OCD has been a lasting theme in an otherwise highly
heterogeneous empirical literature.

CONTEXT: THE ROLE OF ANXIETY

In a recent examination of the current state of classification of psychopa-
thology, Clark et al. (1995) have indicted the heuristic limitations of prevailing
diagnostic nosologies. These authors have noted that the descriptive,
“atheoretical” framework of the DSM, in which disorders are grouped ac-
cording to their most evident shared symptomatology (e.g., anxiety or mood)
has actually begun to impede research into alternative diagnostic criteria (see
also Carson, 1991). The DSM system may serve to supplant theory, with re-
search being guided and constrained by descriptive classifications—originally
instituted for convenience and normative purposes—rather than more mean-
ingful models: “shared phenomenology is only one of many possible ways to
organize psychopathology” (Clark et al., 1995, p. 124). Heterogeneity within
existing diagnostic categories is particularly problematic, as is the separate
categorization of potentially related disorders. Clark et al. have cited OCD as
an example of both of these faults.
Debates over classification have been fueled not only by OCD’s similarity
to quite taxonomically disparate disorders, but also by its dissimilarity to the
syndromes with which it shares a diagnostic category.1 While the DSM system

1
It must be acknowledged that there is considerable diversity in presentation across the anxiety dis-
orders. Indeed, we do not wish to suggest that OCD is alone among the anxiety disorders in having
unique features, or that the DSM system is the only viable option for their classification. However,
the hierarchical model implicit in the current diagnostic framework (American Psychiatric Associ-
ation, 1994)—that their commonality is nosologically more significant than their singularities—has
some empirical support. In factor-analytic studies, models that posit uniqueness of the various anx-
iety disorders as well as a higher-order shared factor have been found to be superior representa-
tions to models that specify their independence and orthogonality only (e.g., Zinbarg & Barlow,
1996; see Spence, 1997 for similar results with a childhood sample). The implications have been
clearly articulated by Brown (1996): “anxiety disorders are related and belong in a common family
or group of disorders . . . findings were consistent with conceptual models positing a trait vulnera-
bility factor that is common to all anxiety disorders” (p. 29). Other findings, such as high levels of
comorbidity among anxiety disorders (Brown & Barlow, 1992), lend themselves to similar interpre-
tation. As indicated elsewhere in this article, similar sources of information show OCD to be an ex-
ception.
582 L. J. SUMMERFELDT AND N. S. ENDLER

has long been lauded for its theoretical neutrality regarding etiology, use of hi-
erarchies and groupings does represent some unarticulated theories about re-
lationships among disorders. As Stein and Hollander (1993) have noted, the
classification of OCD as an anxiety disorder suggests that its linkage with
those syndromes—the presence of anxiety—is stronger than those with other
disorders. In other words, within-group variability is considered less than that
between groups. However, support for this assumption is equivocal. The de-
fining features of the disorder differ qualitatively from those sharing that diag-
nostic category (Insel, Zahn, & Murphy, 1985). These discrepancies are made
more evident, according to Enright and Beech (1990), by treatment failures,
inadequate conceptual models, complex comorbidity patterns, and clinical ob-
servations. These authors have been quite clear on the implications: “OCD is
a distinct disorder qualitatively different from other anxiety disorder catego-
ries” (p. 625).
The role played by anxiety in OCD is unclear. It is not uncommon for pa-
tients to report feelings of irritability, tension, and dissatisfaction—affective
states subsumed in the DSM-IV (American Psychiatric Association, 1994) un-
der the broad term distress (p. 423)—rather than anxiety, per se. This may re-
flect a lack of emotional insight, or the difficulty of clearly defining the term. A
number of theorists have maintained that anxiety is in fact a blend of different
emotions (e.g., Cloninger, 1986; Gray, 1982). Despite this conceptual
vagueness, in clinical descriptions the affective component of such anxiety dis-
orders as specific and social phobias and panic disorder seems quite consistent.
In these cases, anxiety, phenomenologically, is a preparatory, anticipatory
state, definable as “apprehension, tension, or uneasiness from anticipation of
danger, the source of which is largely unknown or unrecognized” (American
Psychiatric Association, 1984, p. 10). In light of this, Barlow (1988, 1991) has
suggested “anxious apprehension” as a more precise descriptive term. The
concept of trait anxiety is used to describe stable individual differences in the
propensity to experience this state (Endler, 1975, 1997).2
This conceptualization seems applicable to other anxiety disorders, where
harm avoidance is clearly a central feature. Its application to the affective ex-
perience associated with OCD is more problematic. The preparatory state de-
scribed by Barlow is not entirely congruent with those most often reported by
OCD sufferers. In many cases, their emotions are not directed toward some

2
Although this article focuses upon anxiety as a clinical characteristic, it must be noted that there
exist two different but overlapping traditions in approaches to anxiety as a psychological construct:
the psychometric and the classificatory. These, which roughly correspond to the distinction be-
tween continuous and categorical perspectives, share several features but have evolved in relative
mutual isolation. The former tradition focuses primarily upon anxiety as a normally distributed in-
dividual differences variable in nonclinical populations, whereas the latter is used to identify homo-
geneous groups of symptoms in clinical populations.
 COGNITIVE BIASES IN OCD 583

tangible future event, but rather seem to reflect a feeling of frustration over a
current perception of being “stuck” (see Leckman, Walker, Goodman,
Pauls, & Cohen, 1994; Rasmussen & Eisen, 1992, 1994). In contrast to phobic
symptoms, rituals in OCD often serve a function less harm-avoidant than ten-
sion-reducing, with a sense of relief gained upon their satisfactory completion
(Leckman et al., 1994; Rasmussen & Eisen, 1992; Tallis, 1996). Unlike other
anxiety syndromes, the individual is often unable to identify an anticipated
threat, a fact recognized even by the existing classificatory framework: “in
some cases individuals perform rigid or stereotyped acts according to idiosyn-
cratically elaborated rules without being able to indicate why they are doing
them” (American Psychiatric Association, 1994, p. 418). In other cases, the fo-
cus of the concern is so prosaic that one is hard-pressed (like the sufferer) to
define it as a harm to be avoided (e.g., a preoccupation with the alignment of
shoelaces, or the chronic need to affirm that patterns in textiles or wall cov-
erings add up to multiples of 2).
In addition to the conflicting self-descriptions offered by patients with
OCD, there is growing empirical evidence suggesting that, while anxiety may
certainly be one among many mood states experienced by those with OCD, it
may not have a comparably primary, and potentially etiological, role as is
played by it in other disorders. Although subjects with OCD have been found
to score highly on self-report measures of anxiety (e.g., Hoehn-Saric, McLeod,
Zimmerli, & Hipsley, 1993), the physiological substrate of these reports has
not been conclusively demonstrated (Benkelfat et al., 1991; Hoehn-Saric,
McLeod, and Hipsley, 1995; Hollander et al., 1991). The implications of this
have been made explicit by Hoehn-Saric et al. (1995): “OCD symptoms trig-
ger anxiety . . . clinically observable anxiety in patients [may be] a conse-
quence rather than the cause of OCD” (Hoehn-Saric et al., 1995, p. 690). This
point echoes others. Following a detailed review of the then-current clinical
and empirical literatures, Reed (1985) was adamant: “there appears to be no
convincing evidence that anxiety plays a significant role in obsessional disor-
ders . . . where it can be identified it seems to be a result rather than a cause of
compulsive activity” (p. 137; see Rachman & Hodgson, 1980 for another inter-
pretation). This is significant, as many models of OCD that grant anxiety a
causal role rest on the assumed predisposing presence of heightened auto-
nomic sensitivity, or hyperarousal.
This secondary role of anxiety clearly differentiates OCD from other anxi-
ety disorders. Not surprisingly, a recent analysis of the structure of anxiety
symptoms in clinical states found loadings of OCD symptoms on both a gen-
eral “trait diathesis” factor, and on discrete anxiety-related factors (i.e., “fear
of fear”) to be among the lowest of all the anxiety disorders (Zinbarg & Bar-
low, 1996). Further evidence of the uniqueness of OCD may be found in re-
cent comorbidity findings. Crino and Andrews (1996) have noted that al-
though comorbid anxiety disorders are often seen in OCD, comorbid OCD in
584 L. J. SUMMERFELDT AND N. S. ENDLER

other anxiety disorders is significantly less common, suggesting the presence

of unique vulnerability factors in its development. Notably, this is implicitly
recognized in the current International Classification of Diseases (ICD-10;
World Health Organization, 1992). Unlike the DSM-IV, OCD is here catego-
rized separately from those syndromes grouped according to the predomi-
nance of anxiety (i.e., phobic or other anxiety disorders). Clearly, the primacy
of anxious apprehension and harm avoidance, so clearly documented in other
anxiety disorders, is more elusive in the case of OCD.

COGNITIVE BIASES AND ANXIETY

The late 1980s and early 1990s have witnessed a surge of interest in the role
of cognitive factors in anxiety, in both clinical and nonclinical populations. A
vast investigative literature has resulted, and a number of comprehensive re-
search programs—using methods adopted from the field of cognitive sci-
ence—have addressed themselves to the specific mechanisms involved. The
work of MacLeod and his colleagues (e.g., MacLeod & Mathews, 1988, 1991;
MacLeod, Mathews, & Tata, 1986) and of Mogg and her collaborators (e.g.,
Mogg, Mathews, Bird, & MacGregor-Morris, 1990; Mogg, Mathews, &
Eysenck, 1992; Mogg et al., 1991; Mogg et al., 1989) are illustrative of these
studies. A review of this literature is beyond the scope of the present article
(the reader is referred to Dalgleish & Watts, 1990; MacLeod, 1991; Ma-
thews & MacLeod, 1994; Williams, Watts, MacLeod, & Mathews, 1997), but
its findings may be summarized in the following way:
1. There is a consistent relationship between clinical anxiety and cognitive
biases, or the facilitated processing of affect-congruent (threatening) in-
formation.
2. This is manifested at an early, relatively automatic (i.e., involuntary; see
McNally, 1995) level of processing; attentional resources are most af-
fected.
3. This automatic bias is evident when the individual is required to allocate
processing priorities to competing alternatives.
4. The personal relevance, or “concern-relatedness” of the threatening
material often accentuates this bias.
5. These tendencies may play a role in both the onset and perpetuation of
clinical anxiety.
Processing biases, then, are associated with the emotional tone, or content,
of information. It appears that underlying anxiety vulnerability is a selective
attentional bias that operates to facilitate the intake of threatening informa-
tion (MacLeod, 1991). Substantial empirical evidence exists for these conclu-
sions (Mineka & Sutton, 1992), as does general agreement about their applica-
bility to clinical anxiety syndromes such as specific and social phobias (e.g.,
 COGNITIVE BIASES IN OCD 585

Watts, McKenna, Sharrock, & Tresize, 1986; Hope, Rapee, Heimberg, &
Dombeck, 1990; Mogg et al., 1992), generalized anxiety disorder (e.g., Mac-
Leod & Mathews, 1991; Mogg et al., 1989), post-traumatic stress disorder
(Foa, Feske, Murdock, Kozak, & McCarthy, 1991; Kaspi, McNally, & Amir,
1995), and panic disorder (e.g., Ehlers, Margraf, Davies, & Roth, 1988). A de-
tailed review of the last decade’s worth of evidence for similar processing bi-
ases in OCD, however, reveals less convincing support.

COGNITIVE BIASES AND OCD

Response Facilitation: Lowered Auditory Thresholds
In one of the first studies in this area, Foa and McNally (1986) used a dich-
otic listening paradigm to determine whether individuals with OCD display a
heightened auditory sensitivity to threatening stimuli. A posttreatment com-
ponent was added to investigate whether this vigilance reflects the fear-rele-
vance of stimuli—presumably attenuated by behavioral treatment—rather
than their mere familiarity. Eleven OCD patients undergoing intensive expo-
sure and response prevention underwent the procedure before and following
treatment. The two taped prose passages used contained 10 occurrences of the
word pick (neutral tape), and, in the threat-relevant tape, a fear-relevant term
selected to represent the dominant concern of each OCD patient (e.g., urine,
radiation). While repeating aloud the passage presented to the dominant ear,
the subject was required to detect target words in both the attended and unat-
tended passage. Both behavioral (i.e., button pressing) and physiological (i.e.,
skin conductance responses) responses to target stimuli were measured. As
hypothesized, pretreatment outcomes on both measures indicated signifi-
cantly greater detection for fear-relevant than for neutral targets in the unat-
tended passage. These significant differences were not observed following
treatment, a result interpreted as further support for the fear hypothesis.
This early study found evidence for enhanced processing of threatening in-
formation in OCD, but the sample used may limit the generalizability of its
findings. Examination of target concerns reported by subjects reveals that 7 of
the 11 (64%) had contamination fears and washing rituals. Although fre-
quently reported, on average neither of these symptoms have been found to
constitute more than 38% of primary OCD symptoms in epidemiologic stud-
ies (see Foa & Kozak, 1995). The overrepresentation of this symptom profile
in Foa and McNally’s (1996) study likely reflected the clinical context in which
subjects were recruited. In a recent review of OCD symptom subtypes in be-
havioral treatment studies, Ball, Baer, and Otto (1996) found that the propor-
tion of individuals with noncleaning symptoms was much lower than is com-
monly reported in epidemiological figures. The authors proposed that this
586 L. J. SUMMERFELDT AND N. S. ENDLER

may reflect the greater tendency of patients with noncleaning symptoms to re-
fuse or drop out of treatment trials, in addition to having poorer treatment re-
sponse (Minichiello, Baer, & Jenike, 1988). As Foa and McNally’s (1986) sub-
jects were all motivated and screened for intensive behavioral treatment, they
may be considered a fairly select group. Notably, presentation of all of their
subjects, including the four with noncleaning symptoms, was phenomenologi-
cally similar to that seen in phobias (e.g., with highly stimulus-specific fears).
This sample’s representativeness of the full OCD population may be question-
able (see Rasmussen & Eisen, 1992, 1994).
One design feature of the study by Foa and McNally (1986) should be
noted. The findings do not actually provide definitive support for the fear hy-
pothesis, as the appropriate analyses to detect posttreatment effects (i.e.,
within-subjects pre-post treatment comparisons) were not conducted (see
Williams et al., 1997). The threat tape repeatedly presented each subject with
a single personally meaningful target word, so results in the pretreatment con-
dition could feasibly have reflected the high degree of familiarity of the idio-
graphic stimuli used. However, as the behavioral treatment entailed increased
exposure to these stimuli, and, therefore, greater familiarity, the lack of post-
treatment effects did indirectly support the fear hypothesis.

Response Disruption: Stroop Paradigm

Although not specifically designed to investigate cognitive biases in OCD,
a study by McNally, Kaspi, Riemann, and Zeitlin (1990) warrants mention.
Using a modified Stroop task, with emotional valence of words as a manipu-
lated variable, these authors sought to identify selective processing of threat
cues in subjects with posttraumatic stress disorder (PTSD). Threat words rele-
vant to OCD (e.g., feces, germs) were used as noncongruent valenced control
stimuli. Although not reported in the body of the study, the authors also ran
10 outpatients with OCD through the protocol (presumably either as controls
or to validate their rationally derived “OCD-relevant” lexical stimuli). The
footnoted results are interesting. The subjects with OCD, who all had contam-
ination obsessions as a primary symptom, did not display significant interfer-
ence on the Stroop task for OCD words.
Subjects with OCD were also included as controls in a investigation by
McNally and his research team of the differential contributions of global emo-
tionality and threat-relevance of words to processing biases in patients with
panic disorder (PD) (McNally, Riemann, Luro, Lukach, & Kim, 1992). A
Stroop paradigm was used, with neutral nonlexical stimuli, and positive, gen-
eral threat (e.g., fearful, anxiety) and threat-specific (e.g., dizzy, choking, col-
lapse, brain tumor) words. Unexpectedly, between-group comparisons were
complicated by the fact that, in response to threat words thought to be panic-
specific, the OCD group displayed Stroop interference similar to that found
 COGNITIVE BIASES IN OCD 587

with the PD group. Furthermore, this interference was greatest for OCDs in
response to bodily sensation terms (e.g., breathless, choking), the group of
words they rated as the least emotional of all the emotionally valenced lexical
stimuli.
As McNally et al. (1992) have noted, all clinical anxiety disorders are asso-
ciated with unpleasant autonomic symptoms. Hence, the threat words used
were not truly unique to panic-related concerns. Moreover, as the authors did
not describe the symptom profiles of the subjects with OCD, it is impossible to
determine the further influence of somatic obsessions (e.g., the fear that one
has developed a serious illness, such as AIDS or cancer). These are not un-
common in OCD (Foa & Kozak, 1995; see also Barsky, 1992), and would cer-
tainly render the bodily sensation terms in the Stroop task more threat-rele-
vant. Nonetheless, despite these points, it should be noted that the results with
the OCD group in the McNally et al. (1992) study are not generally consistent
with the threat-specificity hypothesis. The findings appear even more anoma-
lous when one considers that this sensitivity of subjects with OCD to words
specific to other anxiety disorders was not found in the Stroop study with
PTSD patients previously published by this group (McNally et al., 1990).
The results of a more recent study by McNally and his research team
(McNally et al., 1994) are similarly irregular. Subjects with OCD served as a
clinical control group in a Stroop study of selective processing of personally
meaningful threat words in PD. Neutral, general-threat, and both positive and
threatening panic-relevant words served as stimuli. Contrary to the findings of
McNally et al. (1992), the OCD group did not show greater Stroop interfer-
ence for panic-threat words than neutral ones. Furthermore, and unlike PD
patients, nor did they exhibit interference for general threat words (e.g., infec-
tious, hazard) relative to neutral ones. This latter result is surprising, given
these words’ relevance to contamination obsessions—the most common
symptom in OCD (Foa & Kozak, 1995)—in addition to their general negative
valence. Even more surprising is the finding that subjects with OCD showed a
significantly greater interference for positive (e.g., calm, secure) than for
panic-threat words. These threat words (e.g., dizzy, suffocate, dying) were not
only negatively valenced, but also conceivably relevant to somatic, contamina-
tion, and violence obsessional themes. Clearly, the findings of McNally et al.
(1994), fail to demonstrate selective processing of negatively-valenced mate-
rial in OCD, at either the general or concern-specific level.
A study specifically designed to identify selective processing of threatening
information by subjects with OCD also used a modified Stroop paradigm
(Foa, Ilai, McCarthy, Shoyer, & Murdock, 1993). Responses of two OCD
groups differing in their symptoms (“washers” vs. “nonwashers”) to contami-
nation, general threat, and neutral words and nonwords were compared with
each other, and with those of normal controls. A specificity effect was ex-
pected; contamination words, being more threat-congruent for those with
588 L. J. SUMMERFELDT AND N. S. ENDLER

washing compulsions, were expected to elicit greater color-naming interfer-

ence in that group of subjects. This hypothesis was confirmed, although the
overall statistical effect was found to reflect the significant interference of only
1 of 10 threat-congruent words (unclean). The responses to the general threat
words, however, are worthy of note. The washer subgroup’s interference
scores did not differ significantly from those of either the nonwasher or nor-
mal group. This result is contrary to what one would generally anticipate, in
view of the mixed evidence for absolute specificity of biases in anxiety disor-
ders (see Wells & Matthews, 1994). While concern relatedness should cer-
tainly magnify group differences, all subjects with OCD, including those with
washing compulsions, might be expected to show more selective processing of
generally threatening material than nonclinical subjects. This was true only for
the nonwasher group.
The study by Foa et al. (1993), unlike those reported by McNally and col-
leagues, did find evidence of a processing bias favoring personally threatening
material. This evidence should be interpreted cautiously. Firstly, the effect
was restricted to one word, and was not observed, in the washer subgroup,
with such generally symptom-relevant terms as anxiety, guilt, and stress. As
subjects in the nonwasher group—who did show interference in response to
these general terms—were described as having a blend of OCD symptoms
(e.g., checking, ordering), their results are actually contrary to the threat-spec-
ificity effect found with the other OCD group. Secondly, a measure of state
anxiety, which should have logically been administered at the time of the
Stroop manipulation, was completed by subjects with OCD up to 1 week prior
to the study. Nonclinical controls, by comparison, were administered the mea-
sure immediately prior to manipulation. Thus, there is no way of reliably de-
termining the contributions made by fluctuating anxiety levels. There is em-
pirical evidence for the roles of both state and trait variables in the attentional
biases associated with anxiety (e.g., MacLeod & Rutherford, 1992; Mogg,
Bradley, De Bono, & Painter, 1997), as the authors have acknowledged. Fi-
nally, even if one disregards the fragility of Foa et al.’s findings, it should be re-
membered that the threat-specificity effect was demonstrated with a highly se-
lect subgroup of subjects with OCD. As already mentioned, while
contamination fears and washing compulsions are prevalent symptoms, they
constitute only a portion of the modes of presentation observed in this disor-
der (Foa & Kozak, 1995). The generalizability of the results is, therefore,
limited.

Response Facilitation and Disruption: Dot-Probe Detection

Findings obtained with the Stroop paradigm, used in all but one of the stud-
ies mentioned, may be difficult to interpret. It is possible that observed differ-
ences in color naming—interpreted as a reflection of enhanced attention for
 COGNITIVE BIASES IN OCD 589

threatening stimuli—reflect a response bias, rather than an attentional or per-

ceptual one. Anxiety-disordered subjects may show a tendency to output
emotional words rather than selectively attend to them (see Dalgleish &
Watts, 1990). As Williams et al. (1997) have suggested, disruptions in color
naming may result from time being allocated to the inhibition of this response
tendency, rather than from an allocation of perceptual resources to threaten-
ing material.
In a recent study of attentional biases in OCD, Tata, Leibowitz, Prunty,
Cameron, and Pickering (1996) sought to remedy this problem by adopting a
strategy introduced by MacLeod, Mathews, and Tata (1986). In an effort to
address this general confound in the cognitive bias literature, MacLeod et al.
(1986) used a visual probe paradigm, in which subjects were asked to indicate
as quickly as possible when a small dot on a computer monitor replaced one of
a pair of stimulus words. Anxious subjects appeared to allocate attentional re-
sources toward the location where threat had occurred; this facilitated re-
sponses to probe replacements of threat words and impaired responses when
neutral stimuli were replaced. Control subjects, in contrast, tended to shift at-
tentional resources away from threat stimuli (see also Broadbent & Broad-
bent, 1988). Tata et al. (1996) attempted to replicate this effect with subjects
with OCD, using OCD-relevant lexical stimuli.
In this study, subjects with contamination fears were found to exhibit en-
hanced probe detection when the probe replaced threat-relevant words, but
not nonspecific (social) threat words. Attentional biases in the form of both
vigilance for threatening stimuli were in evidence. This result differentiated
the OCD group clearly from high trait anxious normal controls (who dis-
played biases toward the more general social-threat words), as well as controls
low in trait anxiety. These findings may be seen as evidence for content-spe-
cific attentional biases in OCD, similar to those seen in other anxiety dis-
orders.
The visual probe paradigm certainly addresses the interpretive problems
associated with the Stroop task; but the conclusions of Tata et al. (1996) merit
some other qualifications. Primarily, though the evidence for threat-directed
vigilance biases was robust, there may have been some critical problems with
design sensitivity. Although a total of 57 subjects participated in the study, ex-
clusion of those with outlying response scores resulted in a total of 10 subjects
in the OCD group. Using these data, the investigators performed 11 t-tests
(unadjusted for number of comparisons), 6 or more correlations, and 2 multi-
variate analyses of covariance. Even with the power enhancement provided by
the study’s within-subjects repeated measures design, an adjustment for num-
ber of comparisons might have been recommended.
Some other methodological flaws may be identified in the Tata et al. (1996)
study. High trait anxious controls and subjects with OCD were matched for
590 L. J. SUMMERFELDT AND N. S. ENDLER

depressive symptomatology, on the basis of scores on the Beck Depression In-

ventory (BDI; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961). This mea-
sure, it was reported, was administered “previously.” As the BDI is designed
to be state-specific to a period of 1 week, there is insufficient information to
discern the impact of concurrent depressive symptoms upon these results; this
is not unimportant. Depression is a well-documented concomitant of OCD,
and it has been suggested that depressive affect in OCD may heighten OCD
patients’ general reactivity to affect laden stimuli (Foa et al., 1983). Comorbid-
ity of another sort may also have been influential. The authors did not include
the presence of other anxiety disorders as a variable, or control for it statisti-
cally. Given the prevalent comorbidity in OCD with other anxiety disorders
(Kolada, Bland, & Newman, 1994), and particularly in those with contamina-
tion concerns (Rasmussen & Eisen, 1992), it is not impossible that the pro-
cessing biases observed may have been at least partly determined by a more
global predisposition to clinical anxiety.
Notwithstanding limitations imposed by these issues, elements of the study
by Tata et al. (1996) had considerable methodological rigor (e.g., selection of
lexical stimuli, screening of data, matching of subjects, use of a dot-probe
task). As such, their findings do provide some evidence of attentional biases in
OCD comparable to those found in other anxiety disorders. However, like the
Foa et al. (1993) study, the generalizability of these findings is limited. The au-
thors have made this clear: “(the experiment) used a group of OCD cases se-
lected for their relatively selective focus on issues of contamination . . . caution
must be exercised in extending the present findings to OCD patients other
than those concerned with contamination” (Tata et al., 1996, p. 58).

Response Disruption: Elaborated Stroop Paradigm

The studies reviewed thus far have investigated the specificity of atten-
tional effects in OCD to personally meaningful concerns (e.g., contamina-
tion). A recent study by Lavy, van Oppen, and van den Hout (1994) sought to
elaborate on this. The criticism has been made in the more general attentional
bias literature that commonly utilized threat words are also more emotional
than nonthreat control words, and that conclusions regarding the threat-relat-
edness of bias are consequently suspect (see Martin, Williams, & Clark, 1991;
Mathews & Klug, 1993). In view of this, Lavy et al. (1994) set out, with subjects
with OCD and normal controls, to test three possible hypotheses concerning
the threat-relatedness, emotionality, and concern-relatedness of the stimuli
evocative of attentional biases. Using a Stroop task, the investigators looked
at the responses of subjects with OCD classified as “washers” and “checkers”
(according to their primary compulsions) to words representing combinations
of these three dimensions (e.g., positive and negative checking-relevant
words). Analysis of interference effects revealed that the OCD group showed
 COGNITIVE BIASES IN OCD 591

selective processing only for concern-relevant threatening stimuli. Neither

specificity or emotionality singly predicted attentional biases.
As the authors have noted, this study failed to replicate the “concern-relat-
edness” findings reported with other anxiety disorders (e.g., Mathews & Klug,
1993). However, results do provide partial support for the more general no-
tion that OCD is associated with processing biases favoring threatening stim-
uli. It is worth noting, however, that the two OCD groups were collapsed to-
gether for the purpose of statistical analyses. This may have disguised
meaningful differences, as empirical evidence and clinical observations—
some of it to be discussed later—have suggested that quite different cognitive
idiosyncrasies are found in patients other than those with contamination con-
cerns (see De Silva, 1986; Sher, Mann, & Frost, 1984). Lavy et al. (1994) have
provided insufficient information for the reader to determine whether the
threat-congruent processing effect was primarily determined by the “washer”
groups’ responses; this possibility is not unlikely. Tata et al. (1996) reported
that “cleaning,” but not “checking,” subscales of a measure of OCD symptom-
atology were significantly correlated with attentional bias (i.e., vigilance) in
their study. Relatedly, the lack of robustness of Lavy et al.’s findings may have
reflected heterogeneous response tendencies within the OCD group. Once the
investigators had statistically corrected for the number of comparisons made,
the threat-congruent bias results were only marginally significant.

INTEGRATION
Several conclusions are justified by this review. It appears that support for
cognitive tendencies in OCD analogous to those observed in other anxiety dis-
orders is not robust. The few studies that have been done present problematic
results, both in the form of anomalous or ambiguous findings (e.g., Foa et al.,
1993; McNally et al., 1992; McNally et al., 1994), or in their simple inability to
demonstrate these effects (e.g, McNally et al., 1990). Moreover, studies that
do provide some evidence of attentional biases have exhibited methodological
flaws of varying gravity (e.g., Foa & McNally, 1986; Foa et al., 1993; Tata et al.,
1996), which render determination of the source of these effects difficult. Fi-
nally, even if one chooses to overlook all of these points, in the few cases
where there have been results consistent with the rest of the literature on anxi-
ety and cognitive biases, it has been associated with only one portion of sub-
jects with OCD—those with contamination concerns (e.g., Foa et al., 1993;
Tata et al., 1996, see also Foa & McNally, 1986).
It might be suggested that these results may simply reflect difficulties in
identifying standardized threat-congruent lexical stimuli for OCD subjects.
Individuals with this disorder often report highly circumscribed and peculiar
preoccupations (see McNally et al., 1990). Heterogeneity in the content of ob-
sessional concerns has been named as a chief impediment to this form of re-
search (Obsessive Compulsive Cognition Working Group, 1997). According
592 L. J. SUMMERFELDT AND N. S. ENDLER

to this argument, demonstration of anxiety-related cognitive biases in subjects

with contamination concerns may reflect the thematic consistency of their
fears and the associated ease of identifying concern-relevant words to be used
as experimental stimuli (see Tata et al., 1996).
This methodological explanation has limitations. It should be noted that
there is now some consensus that specific personal relevance of stimuli is not a
sufficient explanation for observed biases in anxiety-disordered samples (Wil-
liams et al., 1997). Apart from this, idiosyncrasies in the content of OC con-
cerns as an explanation for inconsistencies in this literature are not consistent
with clinical observations that the apparent diversity of OC symptoms is de-
ceptive, and belies a limited number of basic stereotypic themes (see Rasmus-
sen & Eisen, 1992). Finiteness and coherence of OC-related concerns has also
been empirically demonstrated. In a factor analytic study of cues predictive of
exacerbation of symptoms in subjects diagnosed with OCD, Ristvedt, Mac-
Kenzie, and Christenson (1993) reported that 75 environmental cues found to
be widely endorsed by subjects could be summarized by only four basic fac-
tors. Such findings suggest that identification of concern-relevant lexical stim-
uli should be fairly straightforward. Finally, although the methodological ex-
planation might be considered adequate for when anxiety-related biases are
not observed at all, it fails to account for such irregular findings as those re-
ported by McNally et al. (1992). Subjects with OCD in this study displayed
cognitive biases in response to lexical terms that were neither concern-rele-
vant nor rated as highly emotional by subjects, a finding at odds with current
ideas about the causes of these effects (see Williams et al., 1997). In light of
such results, McNally et al. (1990), who publish extensively in this field, have
made a suggestion with considerable implications: “unlike other anxiety-dis-
ordered patients, those with OCD may not selectively process threat cues” (p.
400). As we have seen, the one exception may be that subgroup exhibiting
contamination-related symptoms.
This last point is a significant one. Increasingly, allusions to the possibility
that OCD may not be a homogeneous disorder have begun to emerge in the
literature. Much of this is derived from clinical observation (e.g., Rasmus-
sen & Eisen, 1992), although more systematic empirical evidence may be
found in other sources. These include factor analytically derived models of
symptom structure and their associations with other clinical disorders (Baer,
1994; Leckman et al., 1997), the differential relationships of spectrum (i.e., tic-
related) disorders to specific OCD symptom clusters (Leckman et al., 1995)
and to qualitatively distinct modes of subjective experience of OCD symp-
toms (Leckman et al., 1994), and the underrepresentation of certain (i.e., non-
cleaning) symptom subtypes in treatment outcome studies (Ball et al., 1996;
see also Emmelkamp, Kloek, & Blaauw, 1992). These findings, and others,
point to the likelihood that while some forms of OCD may share considerable
phenomenological features with other anxiety syndromes, such as the primacy
 COGNITIVE BIASES IN OCD 593

of high-risk assessment, the presence of identifiable feared stimuli, and harm-

avoidant behaviors, other forms may be quite unique (see Rasmussen, 1994;
Rasmussen & Eisen, 1992, 1994).
This is also likely to be true of cognitive concomitants to the disorder. In re-
cent years, a body of literature on cognitive factors in OCD has been growing
in relative isolation from the work on cognitive biases. The two have had little
mutual influence. The reason for this insularity may be found in their qualita-
tively different emphases. While the latter places paramount emphasis on the
content of obsessional cognition (i.e., its emotional valence and pertinence to
symptom themes), the former body of work has focused on anomalies in the
form of obsessional thinking in clinical and subclinical analogue samples. Ex-
amples include anomalies, deficits, and idiosyncrasies in such domains as logi-
cal reasoning and inference processes (O’Connor & Robillard, 1995), set-
shifting flexibility (Goodwin & Sher, 1992), cognitive inhibitory mechanisms
(Enright & Beech, 1993a, 1993b; see also Wilhelm, McNally, Baer, & Florin,
1996), visuoconstructional capacity (Hollander et al., 1993), confidence in the
accuracy of memories (Constans, Foa, Franklin, & Mathews, 1995; MacDon-
ald, Antony, MacLeod, & Richter, 1997; McNally & Kohlbeck, 1993), deci-
sion-making (Frost & Shows, 1993), reality monitoring, or the personalization
of experience (McNally & Kohlbeck, 1993; Rubenstein, Peynircioglu, Cham-
bless, & Pigott, 1993), and concept formation and categorization (Frost, La-
hart, Dugas, & Sher, 1988; Persons & Foa, 1984; see also Frost & Hartl, 1996).
In many cases, these processing tendencies have been exclusively associated
with symptom profiles not involving contamination concerns—commonly
identified as checking, hoarding, and symmetry/obsessional slowness sub-
types.

SUMMARY
This last basic point may help shed light on the findings in the literature on
OCD and attentional biases. This small body of work is fraught with defects,
anomalies, and inconsistencies. These findings contrast dramatically with the
robust results obtained with other clinical anxiety disorders. When biases are
in evidence it tends to be with a select group of subjects: those with contamina-
tion concerns. Given this, it is not inconceivable that only this subtype of
OCD—or some core characteristic of which it is the most pure manifesta-
tion—may be associated with cognitive tendencies comparable to those found
in other anxiety disorders (i.e., biases at the attentional level associated with
the emotional tone, or content, of information). Cognitive tendencies evident
in other subtypes likely require quite different conceptual frameworks, per-
haps emphasizing aberrations in the form of obsessional cognition indepen-
dent of the content of material being processed. When treated as a unitary di-
agnostic entity, OCD, cognitively, simply does not resemble other anxiety
disorders.
594 L. J. SUMMERFELDT AND N. S. ENDLER

REFERENCES
Alarçon, R. D., Libb, J. W., Boll, T. J. (1994). Neuropsychological testing in obsessive-compulsive
disorder: A clinical review. Journal of Neuropsychiatry, 6, 217–228.
American Psychiatric Association. (1952). Diagnostic and statistical manual of mental disorders.
Washington, DC: Author.
American Psychiatric Association. (1984). Psychiatric glossary (trade ed.). Washington, DC: Au-
thor.
American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders
(4th ed.). Washington, DC: Author.
Baer, L. (1994). Factor analysis of symptom subtypes of obsessive compulsive disorder and their re-
lation to personality and tic disorders. Journal of Clinical Psychiatry, 55, 18–23.
Ball, S. G., Baer, L., & Otto, M. W. (1996). Symptom subtypes of obsessive-compulsive disorder in
behavioral treatment studies: A quantitative review. Behaviour Research and Therapy, 47,
47–51.
Barlow, D. H. (1988). Anxiety and its disorders: The nature and treatment of anxiety and panic. New
York: Guilford Press.
Barlow, D. H. (1991). Disorders of emotion. Psychological Inquiry, 2, 58–71.
Barsky, A. J. (1992). Hypochondriasis and obsessive compulsive disorder. The Psychiatric Clinics
of North America, 15, 791–802.
Beck, A. T., Ward, C. M., Mendelson, M., Mock, J. E., & Erbaugh, J. K. (1961). An inventory for
measuring depression. Archives of General Psychiatry, 4, 561–571.
Benkelfat, C., Mefford, I. N., Masters, C. F., Nordhal, T. E., King, A. C., Cohen, R. M., & Murphy,
O. L. (1991). Plasma catecholamines and their metabolites in obsessive-compulsive disorder.
Psychiatry Research, 37, 321–331.
Berrios, G. E. (1989). Obsessive-compulsive disorder: Its conceptual history in France during the
19th century. Comprehensive Psychiatry, 30, 283–295.
Broadbent, D., & Broadbent, M. (1988). Anxiety and attentional bias: State and trait. Cognition
and Emotion, 2, 235–246.
Brown, T. A. (1996). Validity of the DSM-III-R and DSM-IV classification systems for anxiety dis-
orders. In R.M. Rapee (Ed.), Current controversies in the anxiety disorders (pp. 21-45). New
York: Guilford Press.
Brown, T. A., & Barlow, D. H. (1992). Comorbidity among anxiety disorders: Implications for
treatment and DSM-IV. Journal of Consulting and Clinical Psychology, 60, 835–844.
Cameron, O. G., Thyer, B. A., Nesse, R. M., & Curtis, G. C. (1986). Symptom profiles of patients
with DSM-III anxiety disorders. American Journal of Psychiatry, 143, 1132–1137.
Carson, R. C. (1991). Dilemmas in the pathway of the DSM-IV. Journal of Abnormal Psychology,
100, 302–307.
Clark, L. A., Watson, D., & Reynolds, S. (1995). Diagnosis and classification of psychopathology:
Challenges to the current system and future directions. Annual Review of Psychology, 46, 121–
153.
Cloninger, C. R. (1986). A unified biosocial theory of personality and its role in the development of
anxiety states. Psychiatric Development, 3, 167–226.
Constans, J. I., Foa, E. G., Franklin, M. E., & Mathews, A. (1995). Memory for actual and imagined
events in OC checkers. Behaviour Research and Therapy, 33, 665–671.
Crino, R. D., & Andrews, G. (1996). Obsessive-compulsive disorder and Axis I comorbidity. Jour-
nal of Anxiety Disorders, 10, 37–46.
Dalgleish, T., & Watts, F. N. (1990). Biases of attention and memory in disorders of anxiety and de-
pression. Clinical Psychology Review, 10, 599–604.
De Silva, P. (1986). Obsessional-compulsive imagery. Behaviour Research and Therapy, 24, 333–
350.
 COGNITIVE BIASES IN OCD 595

Dolberg, O. T., Iancu, I., Sasson, Y., & Zohar, J. (1996). The pathogenesis and treatment of obses-
sive-compulsive disorder. Clinical Neuropharmacology, 19, 129–147.
Ehlers, A., Margraf, J., Davies, S., & Roth, W. T. (1988). Selective processing of threat cues in sub-
jects with panic attacks. Cognition and Emotion, 2, 201–220.
Emmelkamp, P. M. G., Kloek, J., & Blaauw, E. (1992). Obsessive compulsive disorders. In P. H.
Wilson (Ed.), Principles and practice of relapse prevention (pp. 213–234). New York: Guilford
Press.
Endler, N. S. (1975). A person-situation interaction model of anxiety. In C. D. Spielberger & I. G.
Sarason (Eds.), Stress and anxiety (Vol. 1, pp. 145–164). Washington, DC: Hemisphere.
Endler, N. S. (1997). Stress, anxiety and coping: The multidimensional interaction model. Canadian
Psychology, 38, 136–153.
Enright, S. J. (1996). Obsessive-compulsive disorder: Anxiety disorder or schizotype? In R. M.
Rapee (Ed.), Current controversies in the anxiety disorders (pp. 161–190). New York: Guilford
Press.
Enright, S. J., & Beech, A. R. (1993a). Reduced cognitive inhibition in obsessive-compulsive disor-
der. British Journal of Clinical Psychology, 32, 67–74.
Enright, S. J., & Beech, A. R. (1993b). Further evidence of reduced cognitive inhibition in obses-
sive-compulsive disorder. Personality and Individual Differences, 14, 387–395.
Enright, S. J., & Beech, A. R. (1990). Obsessional states: Anxiety disorders or schizotypes. An in-
formation processing and personality assessment. Psychological Medicine, 20, 621–627.
Foa, E. B., Feske, U., Murdock, T. B., Kozak, M. J., & McCarthy, P. R. (1991). Processing of threat-
related information in rape victims. Journal of Abnormal Psychology, 100, 156-162.
Foa, E. B., Grayson, E. B., Steketee, G. S., Dopett, G. S., Turner, R. M., & Latimer, P. R. (1983).
Success and failure in the behavioral treatment of obsessive-compulsives. Journal of Consulting
and Clinical Psychology, 51, 287–297.
Foa, E. B., Ilai, D., McCarthy, P. R., Shoyer, B., & Murdock, T. (1993). Information processing in
obsessive-compulsive disorder. Cognitive Therapy and Research, 17, 173–189.
Foa, E. B., & Kozak, M. J. (1995). DSM-IV field trial: Obsessive-compulsive disorder. American
Journal of Psychiatry, 152, 90–96.
Foa, E. B., & McNally, R. J. (1986). Sensitivity to feared stimuli in obsessive-compulsives: A dich-
otic listening analysis. Cognitive Therapy and Research, 10, 477–485.
Frost, R. O., & Hartl, T. L. (1996). A cognitive-behavioral model of compulsive hoarding. Behav-
iour Research and Therapy, 34, 341–350.
Frost, R. O., Lahart, C. M., Dugas, K. M., & Sher, K. J. (1988). Information processing among non-
clinical compulsives. Behaviour Research and Therapy, 26, 275–277.
Frost, R. O., & Shows, D. L. (1993). The nature and measurement of compulsive indecisiveness.
Behaviour Research and Therapy, 31, 683–692.
Goodwin, A. H., & Sher, K. J. (1992). Deficits in set-shifting ability in nonclinical compulsive
checkers. Journal of Psychopathology and Behavioral Assessment, 14, 81–91.
Gray, J. A. (1982). The neuropsychology of anxiety: An enquiry into the functions of the septo-hip-
pocampal system. Oxford, England: Oxford University Press.
Hoehn-Saric, R., McLeod, D. R., Hipsley, P. (1995). Is hyperarousal essential to obsessive-compul-
sive disorder? Archives of General Psychiatry, 52, 688–693.
Hoehn-Saric, R., McLeod, D. R., Zimmerli, W. D., Hipsley, P. A. (1993). Symptoms and physiolog-
ical manifestations in obsessive-compulsive patients before and after treatment with clomi-
pramine. Journal of Clinical Psychiatry, 54, 272–276.
Hollander, E. (1993a). Introduction. In E. Hollander (Ed.), Obsessive-compulsive related disorders
(pp. 1-16). Washington, DC: American Psychiatric Association Press.
Hollander, E. (Ed). (1993b). Obsessive-compulsive related disorders. Washington, DC: American
Psychiatric Association Press.
Hollander, E. (1993c). Obsessive-compulsive spectrum disorders: An overview. Psychiatric Annals,
23, 355–358.
596 L. J. SUMMERFELDT AND N. S. ENDLER

Hollander, E., DeCaria, C., Nitescu, A., Cooper, T., Stover, B., Gully, R., Klein, D. F., & Liebow-
itz, M. R. (1991). Noradrenergic function in obsessive-compulsive disorder: Behavioural and
neuroendocrine responses to clonidine and comparison to healthy controls. Psychiatry Re-
search, 37, 161–177.
Hope, D. A., Rapee, R. M., Heimberg, R. G., & Dombeck, M. J. (1990). Representation of the self
in social phobia: Vulnerability to social threat. Cognitive Therapy and Research, 14, 177–189.
Insel, T. R., Zahn, T., & Murphy, D. L. (1985). Obsessive-compulsive disorder: An anxiety disor-
der? In A. T. Tuma & J. Maser (Eds.), Anxiety and the anxiety disorders (pp. 577–589). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Janet, P. (1903). Les obsessions et la psychasthénie (Vols. 1 and 2, 2nd ed.). Paris: Alcan.
Kaspi, S. P., McNally, R. J., & Amir, N. (1995). Cognitive processing or emotional information in
posttraumatic stress disorder. Cognitive Therapy and Research, 19, 433–444.
Kolada, J. L., Bland, R. C., & Newman, S. C. (1994). Obsessive-compulsive disorder. Acta Psychiat-
rica Scandinavica, 89, 24–35.
Lavy, E., van Oppen, P., van den Hout, M.N. (1994). Selective processing of emotional information
in obsessive compulsive disorder. Behaviour Research and Therapy, 32, 243–246.
Leckman, J. F., Grice, D. E., Boardman, J., Zhang, H., Vitale, A., Bondi, C., Alsobrook, J.,
Peterson, B. S., Cohen, D. J., Rasmussen, S. A., Goodman, W. K., McDougle, C. J., & Pauls,
D. L. (1997). Symptoms of obsessive-compulsive disorder. American Journal of Psychiatry, 154,
911–917.
Leckman, J. F., Walker, D. E., Goodman, W. K., Pauls, D. L., & Cohen, D. J. (1994). “Just right”
perceptions associated with compulsive behaviour in Tourette’s syndrome. American Journal
of Psychiatry, 151, 675–680.
Leckman, J. F., Brice, D. E., Barr, L. C., deVries, A. L. C., Martin, C., Cohen, D. J., McDougle,
C. J., Goodman, W. K., & Rasmussen, S. A. (1995). Tic-related vs. non-tic-related obsessive
compulsive disorder. Anxiety, 1, 208–215.
MacDonald, P. A., Antony, M. M., MacLeod, C. M., & Richter, M. A. (1997). Memory and confi-
dence in memory judgments among individuals with obsessive compulsive disorder and non-
clinical controls. Behaviour Research and Therapy, 35, 497–505.
MacLeod, C. (1991). Clinical anxiety and the selective encoding of threatening information. Inter-
national Review of Psychiatry, 3, 279–292.
MacLeod, C., & Mathews, A. (1988). Anxiety and the allocation of attention to threat. Quarterly
Journal of Experimental Psychology: Human Experimental Psychology, 38, 659–670.
MacLeod, C., & Mathews, A. (1991). Biased cognitive operations in anxiety: Accessibility of infor-
mation or assignment of processing priorities. Behaviour Research and Therapy, 29, 599–610.
MacLeod, C., Mathews, A., & Tata, P. (1986). Attentional bias in emotional disorder. Journal of
Abnormal Psychology, 95, 15–20.
MacLeod, C., & Rutherford, E. M. (1992). Anxiety and the selective processing of emotional infor-
mation: Mediating roles of awareness, trait and state variables, and personal relevance of stimu-
lus materials. Behaviour Research and Therapy, 30, 479–491.
Martin, M., Williams, R. M., & Clark, D. M. (1991). Does anxiety lead to selective processing of
threat-related information? Behaviour Research and Therapy, 29, 147–160.
Mathews, A., & Klug, F. (1993). Emotionality and interference with color-naming in anxiety. Be-
haviour Research and Therapy, 31, 57–62.
Mathews, A., & MacLeod, C. (1994). Cognitive approaches to emotion and emotional disorders.
Annual Review of Psychology, 45, 25–50.
Mavisakkalian, M. R., Hamann, M. S., Haidar, S. A., & deGroot, C. M. (1993). DSM-III personal-
ity disorders in generalized anxiety, panic/agoraphobia, and obsessive-compulsive disorders.
Comprehensive Psychiatry, 34, 243–248.
McNally, R. J. (1995). Automaticity and the anxiety disorders. Behaviour Research and Therapy,
33, 747–754.
 COGNITIVE BIASES IN OCD 597

McNally, R. J., Amir, N., Louro, C. E., Lukach, B. M., Riemann, B. C., & Calamari, J. E. (1994).
Cognitive processing of idiographic emotional information in panic disorder. Behaviour Re-
search and Therapy, 32, 119–122.
McNally, R. J., Kaspi, S. P., Riemann, B. C., & Zeitlin, S. B. (1990). Selective processing of threat
cues in panic disorder. Behaviour Research and Therapy, 28, 407–412.
McNally, R. J., & Kohlbeck, P. A. (1993). Reality monitoring in obsessive-compulsive disorder. Be-
haviour Research and Therapy, 31, 249–253.
McNally, R. J., Riemann, B. C., Luro, C. E., Lukach, B. M., & Kim, E. (1992). Cognitive processing
of emotional information in panic disorder. Behaviour Research and Therapy, 30, 143–149.
Mineka, S., & Sutton, S. K. (1992). Cognitive biases and the emotional disorders. Psychological Sci-
ence, 3, 65-69.
Minichiello, W. E., Baer, L., & Jenike, M. A. (1988). Behavior therapy for the treatment of obses-
sive-compulsive disorder: Theory and practice. Comprehensive Psychiatry, 29, 123–137.
Mogg, K., Bradley, B. P., De Bono, J., & Painter, M. (1997). Time course of attentional bias for
threat information in non-clinical anxiety. Behaviour Research and Therapy, 35, 297–303.
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, 1230–1237.
Mogg, K., Mathews, A., & Eysenck, M. (1992). Attentional bias to threat in clinical anxiety states.
Cognition and Emotion, 5, 221–238.
Mogg, K., Mathews, A., Eysenck, M., & May, J. (1991). Biased cognitive operations in anxiety: Ar-
tefact, processing priorities, or attentional search? Behaviour Research and Therapy, 29, 459–
467.
Mogg, K., Mathews, A., & Weinman, J. (1989). Selective processing of threat cues in anxiety states:
A replication. Behaviour Research and Therapy, 27, 317–323.
Obsessive Compulsive Cognitions Working Group. (1997). Cognitive assessment of obsessive-
compulsive disorder. Behaviour Research and Therapy, 35, 667–681.
O’Connor, K., & Robillard, S. (1995). Inference processes in obsessive-compulsive disorder: Some
clinical observations. Behaviour Research and Therapy, 33, 887–896.
Oldham, J. M., Hollander, E., & Skodol, A. E. (Eds.). (1996). Impulsivity and compulsivity. Wash-
ington, DC: American Psychiatric Press.
Pauls, D. L. (1992). The genetics of OCD and Gilles de la Tourette’s Syndrome. Psychiatric Clinics
of North America, 15, 759–766.
Persons, J. B., & Foa, E. B. (1984). Processing of fearful and neutral information by obsessive-com-
pulsives. Behavior Research and Therapy, 22, 259–265.
Rachman, S. (1997). A cognitive theory of obsessions. Behaviour Research and Therapy, 35, 793–
802.
Rachman, S., & Hodgson, R. J. (1980). Obsessions and compulsions. Englewood Cliffs, NJ: Pren-
tice-Hall.
Rasmussen, S. A. (1994). Obsessive compulsive spectrum disorders. Journal of Clinical Psychiatry,
55, 89–91.
Rasmussen, S. A., & Eisen, J. L. (1992). The epidemiology and clinical features of obsessive com-
pulsive disorder. The Psychiatric Clinics of North America, 15, 743–758.
Rasmussen, S. A., & Eisen, J. L. (1994). Clinical features and phenomenology of obsessive-compul-
sive disorder. Psychiatric Annals, 19, 67–73.
Reed, G .F. (1985). Obsessional experience and compulsive behaviour: A cognitive-structural ap-
proach. Toronto: Academic Press.
Reed, G. F. (1991). The cognitive characteristics of obsessional disorder. In P. A. Magaro (Ed.),
Cognitive bases of mental disorders (pp. 77—99). London: Sage.
Ristvedt, S. L., MacKenzie, T. B., & Christenson, G. A. (1993). Cues to obsessive-compulsive
symptoms: Relationships with other patient characteristics. Behaviour Research and Therapy,
31, 721–729.
598 L. J. SUMMERFELDT AND N. S. ENDLER

Rubenstein, C. S., Peynircioglu, Z. F., Chambless, D. L., & Pigott, T. A. (1993). Memory in sub-
clinical obsessive-compulsive checkers. Behaviour Research and Therapy, 31, 759–765.
Salkovskis, P. M. (1985). Obsessional-compulsive problems: A cognitive-behavioural analysis. Be-
haviour Research and Therapy, 31, 759–765.
Salkovskis, P. M. (1989). Cognitive-behavioural factors and the persistence of intrusive thoughts in
obsessional problems. Behaviour Research and Therapy, 27, 677–682.
Sher, K., Mann, B., & Frost, R. (1984). Cognitive dysfunction in compulsive checkers: Further ex-
plorations. Behaviour Research and Therapy, 22, 493–502.
Spence, S. H. (1997). Structure of anxiety symptoms among children: A confirmatory factor ana-
lytic study. Journal of Abnormal Psychology, 106, 280–297.
Stein, D. J., & Hollander, E. (1993). The spectrum of obsessive-compulsive-related disorders. In E.
Hollander (Ed.), Obsessive compulsive related disorders (pp. 241–271). Washington, DC:
American Psychiatric Press.
Stein, D. J., Hollander, E., & Skodol, A. E. (1993). Anxiety disorders and personality: A review.
Journal of Personality Disorders, 7, 87–104.
Summerfeldt, L. J., Huta, V. M., & Swinson, R. P. (1998). Personality and obsessive compulsive dis-
order. In R. P. Swinson, M. Antony, S. Rachman, & M. Richter (Eds.), Obsessive compulsive
disorder: Theory research and treatment (pp. 79–119). New York: Guilford Press.
Tallis, F. (1996). Compulsive washing in the absence of phobic and illness anxiety. Behaviour Re-
search and Therapy, 34, 361–362.
Tata, P. R., Leibowitz, J. A., Prunty, M. J., Cameron, M., & Pickering, A. D. (1996). Attentional
bias in obsessional compulsive disorder. Behaviour Research and Therapy, 34, 53–60.
Tynes, L. L., White, K., & Steketee, G. S. (1990). Toward a new nosology of obsessive compulsive
disorder. Comprehensive Psychiatry, 31, 465–480.
Watts, F. N., McKenna, F. P., Sharrock, R., & Tresize, L. (1986). Colour naming of phobia-related
words. British Journal of Clinical Psychology, 77, 97–108.
Wells, A., & Matthews, G. (1994). Attention and emotion: A clinical perspective. Hillsdale, NJ: Law-
rence Erlbaum Associates.
Westphal, C. (1877). Uber Zwangsvorstellungen. Berliner Klinische Wochenschrift, 46, 669–684.
Wilhelm, S., McNally, R. J., Baer, L., & Florin, I. (1996). Directed forgetting in obsessive-compul-
sive disorder. Behaviour Research and Therapy, 34, 633–641.
Williams, J. M. G., Watts, F. N., MacLeod, C., & Mathews, A. (1997). Cognitive psychology and
emotional disorders (2nd ed.). Toronto: Wiley.
World Health Organization. (1992). The ICD-I0 classification of mental and behavioral disorders:
Clinical descriptions and diagnostic guidelines. Geneva: Author.
Zinbarg, R. E., & Barlow, D. H. (1996). Structure of anxiety and the anxiety disorders: A hierarchi-
cal model. Journal of Abnormal Psychology, 105, 181–193.