FDFHFDH

COGNITIVE NEUROPSYCHIATRY
2007, 12 (1), 46 77
Hopping, skipping or jumping to conclusions? Clarifying

the role of the JTC bias in delusions
Cordelia Fine
Centre for Applied Philosophy and Public Ethics, Department of Philosophy,
University of Melbourne, Victoria, Australia
Mark Gardner
Department of Psychology, University of Westminster, London, UK
Jillian Craigie
School of Philosophy and Bioethics, Monash University, Clayton, Victoria,
Australia
Ian Gold
School of Philosophy and Bioethics, Monash University, Clayton, Victoria,
Australia, and Departments of Philosophy & Psychiatry, McGill University,
Montreal, Quebec, Canada
Introduction . There is substantial evidence that patients with delusions exhibit a

reasoning bias *known as the ‘‘jumping to conclusions’’ (JTC) bias *which leads
them to accept hypotheses as correct on the basis of less evidence than controls. We
address three questions concerning the JTC bias that require clarification. Firstly,
what is the best measure of the JTC bias? Second, is the JTC bias correlated
specifically with delusions, or only with the symptomatology of schizophrenia? And
third, is the bias enhanced by emotionally salient material?
Methods. To address these questions, we conducted a series of meta-analyses of
studies that used the Beads task to compare the probabilistic reasoning styles of
individuals with and without delusions.
Correspondence should be addressed to Cordelia Fine, Centre for Applied Philosophy and
Public Ethics, Department of Philosophy, University of Melbourne, Victoria 3010, Australia.
E-mail: cfine@unimelb.edu.au
This work was supported by an Australian Research Council Large Grant (no. A00103679), an
Australian Research Council Discovery-Projects Grant (no. 0343348), and a Senior Fellowship to
IG in the McDonnell Project in Philosophy and the Neurosciences, Simon Fraser University. We
would also like to thank our reviewers *in particular, Emmanuelle Peters *for many helpful
comments on earlier versions of the paper.
# 2007 Psychology Press, an imprint of the Taylor & Francis Group, an informa business
http://www.psypress.com/cogneuropsychiatry DOI: 10.1080/13546800600750597
JTC BIAS AND DELUSIONS 47
Results. We found that only one of four measures of the JTC bias *‘‘draws to
decision’’ *reached significance. The JTC bias exhibited by delusional subjects *
as measured by draws to decision *did not appear to be solely an epiphenomenal
effect of schizophrenic symptomatology, and was not amplified by emotionally
salient material.
Conclusions. A tendency to gather less evidence in the Beads task is reliably
associated with the presence of delusional symptomatology. In contrast, certainty
on the task, and responses to contradictory evidence, do not discriminate well
between those with and without delusions. The implications for the underlying basis
of the JTC bias, and its role in the formation and maintenance of delusions, are
discussed.
INTRODUCTION: THE ‘‘JUMPING TO CONCLUSIONS’’ BIAS

Delusions are currently defined as ‘‘belief[s] based on incorrect inference’’
(American Psychiatric Association, 1994, p. 765). Accordingly, one sig-
nificant approach to the explanation of delusion hypothesises that a
reasoning bias contributes to delusion formation and/or retention. The
seminal articulation of this hypothesis is that of Garety and colleagues (e.g.,
Garety & Freeman, 1999; Garety & Hemsley, 1994) who have proposed that
at least some individuals with delusions possess a reasoning style char-
acterised by ‘‘a tendency or bias to the early acceptance and, to a lesser
extent, the early rejection of hypotheses’’ (Garety & Freeman, 1999, p. 127).
Such a reasoning style has been encapsulated in the phrase ‘‘jumping to
conclusions’’ (hereafter ‘‘JTC’’). A JTC style is generally considered to be
only one of a number of factors that potentially contribute in complex ways
to the formation and retention of delusions (e.g., Garety, Hemsley, &
Wessely, 1991; Garety, Kuipers, Fowler, Freeman, & Bebbington, 2001).
Other contributors may include personality, affective, motivational, and
perceptual factors, the relative importance of which will vary from patient to
patient. As Garety and Freeman (1999) note, the nature of such models
makes it difficult to generate clear predictions about group differences.
Nonetheless, a substantial body of research has found evidence of a JTC bias
in patients with delusions. In a review of the JTC literature, Garety and
Freeman (p. 147) concluded that there is ‘‘strong support for a reasoning
bias in people with delusions which is best described as a data-gathering bias,
a tendency for people with delusions to gather less evidence than controls so
that they jump to conclusions’’.
Whether a JTC bias plays a causal role in delusion formation remains
contentious. While suitably cautious, Garety and Freeman (1999, p. 127)
speculate in favour of a causal relation. They suggest that a JTC reasoning
style may ‘‘under certain conditions, contribute to erroneous inferences and,
therefore, to delusion formation’’. Other researchers concur (e.g., Colbert &
Peters, 2002; Dudley, John, Young, & Over, 1997a; Linney, Peters, & Ayton,
48 FINE ET AL.
1998; McGuire, Junginger, Adams, Burright, & Donovick, 2001; Peters &
Garety, 2006; van Dael, Vermissen, Janssen, Myin-Germeys, van Os, &
Krabbendam, 2006). A JTC bias has also been suggested as a possible
contributory factor in several well-known multifactorial accounts of the
formation of specific categories of delusion, for example, persecutory
delusions (e.g., Bentall, Corcoran, Howard, Blackwood, & Kinderman,
2001) and the Capgras delusion (e.g., Young, 2000). The claim that a
reasoning bias contributes to the formation of delusions has also had an
impact on the development of cognitive therapies for delusions, some of
which now target the reasoning processes thought to be involved in
delusional belief formation and the maintenance of those beliefs (see Rector
& Beck, 2002).
WHAT IS THE JTC BIAS?

Dependent variables used to measure JTC
Investigation of a JTC bias in patients with delusions has been primarily
conducted within a Bayesian reasoning framework in which participants
assess the probabilities of events on the basis of empirical evidence. In the
paradigmatic experimental task *the Beads task *participants are told that
beads will be drawn from one of two jars. The jars contain beads of two
different colours, in complementary ratios, for example, 85:15 red to green,
versus 85:15 green to red. (In various manipulations of the task, different
stimulus pairs are substituted for the two bead colours, or the ratios of the
two types of stimuli are changed.) Participants are shown the beads
supposedly being drawn from the jar, the selection of which is actually
predetermined, and are required to guess from which jar they are being
drawn. The central prediction of the task is that if delusional patients do in
fact have a JTC tendency, they will make a guess about the jar more quickly,
or with greater certainty, than controls.
To assess these predictions, two different methodologies have been used.
In the ‘‘draws to decision’’ methodology, participants request as many beads
as they deem necessary to decide from which jar the beads are being drawn.
The prediction is that patients with delusions will require significantly fewer
beads to ‘‘draw to decision’’ than control groups. A small number of studies
have used the related dichotomous variable of ‘‘extreme responding’’, that is,
requiring only one (or sometimes two) beads to make a decision. In the
‘‘draws to certainty’’ methodology, all participants see the same number of
beads, and this number is predetermined by the experimenter. After each
bead draw, participants are asked how certain they are that the beads are
being drawn from a particular jar. The variables of interest are the number of
draws required to reach a high level of certainty, and/or the mean level of
certainty on the early trials of the task. A variant of this procedure is the
‘‘graded estimates’’ approach. Here, after each trial the participant is asked
whether the bead is definitely, almost certainly, or probably from jar A or B,
or whether there is currently no preference.
The ‘‘draws to certainty’’ methodology is also used to assess a second
prediction of the JTC hypothesis, that patients with delusions will be more
likely than controls to reject hypotheses in response to contradictory
evidence (Garety & Hemsley, 1994). In what we refer to as ‘‘response to
potentially disconfirmatory evidence’’, the dependent variable is the change
in certainty in response to a single bead that potentially disconfirms the
participant’s likely hypothesis. For example, Garety et al. (1991) calculated
participants’ certainty after Bead 3 (the third of three pink beads) minus
certainty after Bead 4 (the first green bead), as a measure of ‘‘response to
potentially disconfirmatory evidence’’.
Some researchers have also examined participants’ change in certainty
when, in the latter half of the experiment, the evidence is genuinely
disconfirmatory. Participants are presented with a sequence of beads such
as the one below:
AAABAAAABABBBABBBBAB
In this sequence, the first 10 beads suggest that beads are being drawn from
the ‘‘mostly As’’ jar, while the second 10 beads suggest that the beads are
being drawn from the ‘‘mostly Bs’’ jar. It is then assessed whether patients
with delusion are the same as, or different from, control groups in the way
they lose confidence in their initial hypothesis. We refer to this dependent
variable as ‘‘response to reversal’’.
The range of dependent variables used to tap JTC reflects uncertainty in
the literature regarding what, precisely, are useful and valid measures of
‘‘jumping to conclusions’’. There appears to be consensus that hastiness in
decision making derives from a lower threshold for amount of information
required (measured by the dependent variable ‘‘draws to decision’’), rather
than a bias in reasoning about probabilities per se (e.g., Dudley, John,
Young, & Over, 1997b; Garety & Freeman, 1999; Moritz, Woodward, &
Hausmann, 2006; Peters, Thornton, Siksou, Linney, & MacCabe, 2005).
However, ‘‘draws to certainty’’ has been used as a measure of JTC bias in
several recent studies (e.g., Moritz & Woodward, 2005; Peters & Garety,
2006; Peters et al., 2005). The issue of whether or not ‘‘draws to certainty’’
predicts the presence of delusions is an important one, since the answer
could constrain in significant ways our understanding of the JTC
phenomenon. Finally, it is not clear whether the JTC phenomenon should
also be taken to include a tendency to ‘‘jump to new conclusions’’, indexed
by the dependent variables ‘‘response to potentially disconfirmatory
50 FINE ET AL.
evidence’’ and ‘‘response to reversal’’. After all, one feature of delusional

beliefs is their apparent incorrigibility in the face of contradictory evidence
(e.g., American Psychiatric Association, 1994). One might then expect
patients with delusions to be less responsive to contradictory evidence than
controls (e.g., Moritz & Woodward, 2005). Thus whether or not the JTC bias
observed in patients with delusions is restricted to the initial formation of a
hypothesis, or also extends to the individual’s response to later evidence, is
also an important question.
Theoretical accounts of the JTC bias

Determining which putative measures of JTC are statistically reliable in
distinguishing between those with and without delusions is of critical
importance if we are to refine theoretical accounts of the underlying cause
of the JTC bias. Accounts of the information processing style hypothesised
to be responsible for patients’ tendency to jump to conclusions fall into two
main categories that we will refer to as ‘‘information integration’’ and
‘‘motivation’’ accounts. We discuss them in turn.
The ‘‘information integration’’ account of Menon and colleagues
(Menon, Woodward, Pomarol-Clotet, McKenna, & McCarthy, 2005;
Menon, Pomarol-Clotet, McKenna, & McCarthy, 2006), for example,
extrapolates from the suggestion that in schizophrenia, currently experienced
stimuli are imbued with an abnormal salience (Kapur, 2003). As a
consequence of this abnormal stimulus valuation, patients inappropriately
weight evidence in the Beads task. As a result, the decision-making threshold
is crossed more quickly. In another account of this type, Bentall and Young
(e.g., Young & Bentall, 1995) have proposed that delusional patients have
difficulty making use of sequential information. According to this account,
they make hasty decisions on the Beads task in order to avoid making a
judgement on the basis of a sequence of information.
If the cause of a JTC style is placing too much value on a current piece of
evidence (as proposed by Menon and colleagues), then one should see both
reduced ‘‘draws to decision’’ as well as reduced ‘‘draws to certainty’’ on the
grounds that if a current piece of evidence is overvalued then this should give
rise both to hastiness in judgement as well as an inflated confidence in one’s
hypothesis. In addition, this account predicts that delusional patients will
also ‘‘jump to new conclusions’’ because they overvalue current stimuli
relative to beads that have gone before. Thus confidence in their initial
hypothesis should diminish significantly more than controls following
disconfirmatory evidence. By contrast, it seems to us that a difficulty in
processing sequential information (as suggested by Bentall and colleagues)
predicts hastiness in decision making, but does not imply that delusional
patients will have greater confidence in their hypotheses, or that their

confidence will be unusually influenced by disconfirmatory evidence.
In contrast to ‘‘information integration’’ accounts of the JTC bias,
‘‘motivation’’ accounts hypothesise that patients’ hasty reasoning style arises
from a motivation to confirm beliefs. Bentall et al. (2001), for example, have
suggested that the JTC bias arises from a ‘‘need for closure’’. The need for
closure (NFC; Kruglanski & Webster, 1996) is the ‘‘desire for definite
knowledge on some issue and the eschewal of confusion and ambiguity’’
(p. 278). Individuals with an NFC are thought to consider less evidence
before making a decision, entertain fewer hypotheses, experience greater
certainty about their judgement and cling to it, ‘‘becoming impervious to
subsequent data’’ (p. 265). If the JTC bias arises from an enhanced NFC,
then it should manifest in reduced ‘‘draws to decision’’ and ‘‘draws to
certainty’’, and in greater stability in the level of confidence following
contradictory evidence.
Dudley and Over (2003) have recently suggested that a high NFC in
delusional patients reflects a need for ‘‘threat confirmation’’, and is part of a
‘‘confirmatory reasoning style’’. The JTC bias, in their view, arises from this
confirmatory style of reasoning: ‘‘people without delusions demonstrate a
normal tendency to confirm danger-related conditional statements [but]
people with delusions perceive danger and threat where others do not. Hence,
this normal reasoning style is extended to non-threatening situations’’
(Dudley & Over, 2003, p. 263). This account predicts reduced data gathering
in patients with delusions (since they are hypothesised to be less motivated to
find evidence that disconfirms their hypothesis). However, it is not clear that
Dudley and Over’s position predicts any differences in certainty ratings in
either confirmatory or disconfirmatory phases of the Beads task. Their
position, therefore, predicts that delusional subjects will require fewer ‘‘draws
to decision’’ but not fewer ‘‘draws to certainty’’ compared to controls.
Summary
Theoretical understanding of the JTC bias currently lacks precision.
Different studies have used different dependent measures to tap a JTC
bias and it is not yet clear which measure(s) best discriminate between
subjects with and without delusions. Reflecting this uncertainty, current
accounts of the cause underlying the tendency to jump to conclusions differ
in the predictions they make regarding how the JTC bias should manifest in
patients with delusions. An investigation into which of these measures best
discriminates the reasoning style of delusional patients from controls is
needed therefore to clarify the nature of the putative JTC bias in delusion,
and to indicate viable accounts of an underlying mechanism.
52 FINE ET AL.
STUDY 1: A META-ANALYTIC COMPARISON OF

DEPENDENT MEASURES OF THE JTC BIAS
The aim of Study 1 was therefore to address the question of which
operational definition(s) of the JTC bias best discriminate between
individuals with and without delusions.
‘‘Draws to decision’’ and ‘‘draws to certainty’’ are the most frequently
utilised measures of JTC. In their review, Garety and Freeman (1999)
concluded that ‘‘draws to decision’’ successfully discriminates between the
presence and absence of delusions, but ‘‘draws to certainty’’ does not. To
date, 10 studies have used the ‘‘draws to decision’’ procedure. Six of these
studies demonstrated significant differences between patients with delusions
and psychiatric controls (Dudley et al., 1997a,b; Fear & Healy, 1997; Moritz
& Woodward, 20051; Peters & Garety, 2006; Peters et al., 2005). The
remaining four studies demonstrated significant differences with nonpsy-
chiatric controls (Conway et al., 20022; Garety et al., 1991; Huq, Garety, &
Hemsley, 1988; Menon et al., 2006).3
Six studies have used the ‘‘draws to certainty’’ or similar procedures.
Garety et al. (1991) found no significant differences between patients with
delusions and either psychiatric or nonpsychiatric controls using this
procedure. Peters et al. (2005) found no differences on this measure between
psychiatric patients grouped according to the presence or absence of
delusional symptomatology. Similarly, Moritz and Woodward (2005), using
the ‘‘graded estimates’’ procedure, found no differences between patients
with schizophrenia grouped according to the presence or absence of
delusional symptomatology. Using the same measure, Young and Bentall
(1997) found mixed support for a difference between patients with delusions
and nonpsychiatric controls (using both standard and emotionally salient
versions of the Beads task), but observed no differences in comparisons with
depressed psychiatric controls. Fear and Healy (1997) found a significant
difference between patients with delusions and psychiatric controls with
obsessive-compulsive disorder (OCD). However, the patients with delusions
did not differ from the nonpsychiatric controls, and the patients with OCD
required a significantly greater ‘‘draws to certainty’’ than did nonpsychiatric
controls. This suggests that the significant difference in ‘‘draws to certainty’’
observed between the patients with delusions and the patients with OCD
should perhaps be attributed to increased ‘‘draws to certainty’’ in the latter
1
However, no significant differences were found in comparison with a nondelusional group
with schizophrenia.
2
No psychiatric control group was used.
3
For the harder 60:40 ratio. However, no significant differences were observed in comparison
with a nondelusional group with schizophrenia, or when a memory aid was provided.
group (see Pélissier & O’Connor, 2002; Volans, 1976; also Dudley & Over,
2003), rather than a reduction in the delusional group, relative to normal
nonpsychiatric performance. The only unambiguous support for reduced
‘‘draws to certainty’’ in patients with delusions comes from Peters and
Garety (2006), who found that their delusional group showed a significantly
greater mean certainty over the first three trials than did both psychiatric
and nonpsychiatric controls.
Overall, it appears that ‘‘draws to certainty’’ does not discriminate well
between the presence or absence of delusions once the confounding effects of
psychiatric symptomatology are taken into account. From this it has been
inferred that patients with delusions have a tendency to seek less information
prior to reaching a decision (and thus have a reduced ‘‘draws to decision’’),
rather than a tendency to differ with respect to certainty (Dudley et al.,
1997b; Garety & Freeman, 1999). In line with this conclusion, a number of
studies have demonstrated that patients with delusions show normal
sensitivity to variations in the ratio of beads in the Beads task. Thus when
the ratio of different coloured beads in the jar is changed (for example,
reduced to 60:40) patients with delusions, like controls, appropriately change
the number of beads they require for a decision or for certainty (e.g., Dudley
et al., 1997b; Young & Bentall, 1997).
Five studies have used ‘‘responses to potentially disconfirmatory evi-
dence’’ as a dependent variable. The majority of these found that individuals
with delusions become significantly less certain about their hypotheses
following potentially disconfirmatory evidence in comparison with non-
psychiatric, but not psychiatric, control groups (Fear & Healy, 1997; Garety
et al., 1991; Peters & Garety, 2006; Young & Bentall, 1997). One exception is
Moritz and Woodward (2005), who found that this dependent variable
distinguished between delusional patients with schizophrenia and nondelu-
sional patients with schizophrenia, as well as psychiatric and healthy
controls.
A fourth variable, ‘‘responses to reversal’’, has also been used as a
measure of a JTC style. Overall, the evidence from studies using this variable
also offers little support for the idea that patients with delusions ‘‘jump to
new conclusions’’. Three studies using this measure found no significant
differences between patients with delusions and control groups (Dudley et
al., 1997b; Fear & Healy, 1997; Young & Bentall, 1997). Moritz and
Woodward (2005) found some evidence of greater adjustment of certainty
responses in delusional patients with schizophrenia compared with psychia-
tric and nonpsychiatric controls. However, significant effects were not
observed for all disconfirmatory beads (only two of the eight beads in the
latter half of the experiment that disconfirmed the original hypothesis), and
no significant differences were seen when these patients were compared with
nondelusional patients with schizophrenia. Brankovic and Paunovic (1999)
54 FINE ET AL.
found that delusional patients showed significantly smaller ‘‘responses to

reversal’’, compared with both psychiatric and nonpsychiatric controls, but
this may be due to their unconventional method of quantifying this putative
measure of JTC.
Garety and Freeman (1999, p. 131) have concluded that ‘‘people with
delusions may be more ready to abandon existing hypotheses and form new
ones, again on the basis of little evidence’’. Relating this hypothesised
manifestation of the JTC bias to the clinical setting, Moritz and Woodward
(2005, p. 204) noted that ‘‘many deluded schizophrenic patients dramatically
switch their attitude towards the persecutor or persons embedded in the
paranoid belief’’. However, as noted by Peters and Garety (2006) and
indicated by the data reviewed above, it is not clear that measures of
response to contradictory evidence do in fact often successfully differentiate
between delusional and other psychiatric control groups. The data justify
some suspicion as to whether increased ‘‘response to potentially disconfir-
matory evidence’’ is actually associated with delusions, or merely with
psychiatric symptomatology, and there is little evidence that patients with
delusions show an increased ‘‘response to reversal’’.
A qualitative summary of the research thus suggests that a reduced
number of ‘‘draws to decision’’ may be most strongly associated with the
presence of delusions. The status of the variables ‘‘draws to certainty’’ and
‘‘response to potentially disconfirmatory evidence’’ are less clear, given the
role psychiatric symptomatology appears to play in contributing to these
effects. ‘‘Response to reversal’’ appears to find little support from the
literature.
In order to provide a more objective evaluation of the association of
different JTC measures with delusions, we conducted a meta-analysis. We
were interested in three main questions. First, we investigated which
dependent variable best discriminates between individuals with and without
delusions. Second, we explored the impact of comparison group (psychiatric
control vs. nonpsychiatric control) on effect size to establish whether
psychiatric symptomatology contributes to the JTC effect. Third, in the
case of evidence for an overall contribution from psychiatric symptomatol-
ogy across all measures of JTC, we explored the effect of psychiatric
symptomatology on each of the putative measures of JTC.
Method
Studies to be included in the meta-analysis were identified by carrying out an
electronic search using the Medline and PsycInfo databases. Search terms
used included: delusion, jumping to conclusions, reasoning. In addition, the
review of Garety and Freeman (1999), and the reference lists of other more
recent articles in this area, were used to identify potentially relevant

publications. Finally, a reviewer, E. Peters, provided a number of important
prepublications that were included in our study.
Studies included were those examining the association between clinical
delusion symptomatology and reasoning bias by either (1) comparing a
group of patients with delusions with a control group without delusions or
(2) examining correlations between delusional symptomatology and reason-
ing within a clinical sample. One study, Brankovic and Paunovic (1999), was
excluded on the grounds that highly unconventional measures of JTC were
employed. We also did not include two studies that presented only the
dichotomous dependent variable, ‘‘extreme responding’’ (Garety et al., 2005;
van Dael et al., 2006). After searching the literature and screening for
inclusion criteria, 47 effect sizes were extracted from 12 studies. The
published studies included are summarised in Table 1. Effect sizes were
also extracted from two additional studies that employed a subclinical
‘‘prone to delusions’’ group. These effect sizes are reported in the Results and
Discussion of Study 2.
Effect sizes were calculated primarily from published means and standard
deviations. Where these were not available, effect sizes were derived from
reported statistics such as F, t , or r (see Rosenthal, 1991, for formula). The
effect size estimate, g (Hedges, 1981), refers to the standardised difference
between delusional and nondelusional groups’ performance on a single
measure from the Beads task and is reported for each study in Table 1. In
cases where studies reported repeated testing or effects relating to
manipulations not of concern to the present investigation (e.g., bead ratio,
presence of a memory aid), separate estimates of effect size were first
computed before being combined using the arithmetic mean (Rosenthal,
1991). However, for some studies, more than one effect size was extracted.
We included all available effects from a particular study relating to our
moderators of interest, study design, and dependent variable. Although
caution has been advised regarding the analysis of nonindependent effects
(e.g., Rosenthal, 1991), it has been argued that violation of the assumption
of independence has little effect on statistical precision (Tracz, Elmore, &
Pohlmann, 1992).
The meta-analysis employed procedures described by Rosenthal (1991).
Two types of analysis were used. In order to assess whether the pooled data
from several similar published studies jointly achieved significance, reported
effects were combined using the method of adding Z’s (the Stouffer method;
Mosteller & Bush, 1954). In order to assess whether a moderator variable
had a reliable effect on the results of published studies, reported effects were
compared using focused comparisons (Rosenthal & Rubin, 1982).
56
TABLE 1
Summary of published studies examining the association between clinical delusion symptomatology and reasoning bias included in
FINE ET AL.
the meta-analyses
Design: Groups Effect size (g)
Reference Delusional Control Decision Certainty Pot. Disc. Reversal
Huq et al. (1988) Schizophrenia Psychiatric 0.94

Nonpsychiatric 1.00
Garety et al. (1991) Schizophrenia Psychiatric 0.74 /0.22 0.31
Nonpsychiatric 1.15 0.42 0.19
Delusional disorder Psychiatric 0.27 0.32 0.62
Nonpsychiatric 0.74 1.10 0.35
Mortimer et al. (1996) Schizophreniaa Schizophrenia without delusionsa /0.37
Dudley et al. (1997a) Schizophrenia Psychiatric 0.88
Nonpsychiatric 0.96
Dudley et al. (1997b) Schizophrenia Psychiatric 1.32 0.32
Nonpsychiatric 1.43 0.43
Fear & Healy (1997) Delusional disorder Psychiatricb 0.98 1.72 0.00 /0.04
Nonpsychiatric 0.96 0.18 0.84 0.00
Young & Bentall (1997) Schizophreniac Psychiatric 0.03d 0.13 /0.06
Nonpsychiatric 0.46d 0.82 0.35
Conway et al. (2002) Delusional disorder Nonpsychiatric 0.96
Menon et al. (2006) Schizophrenia Schizophrenia without delusions 0.12
Nonpsychiatric 0.34
Moritz & Woodward (2005) Schizophrenia Schizophrenia without delusions 0.14 /0.22d
Psychiatrice 0.79 0.50d
Nonpsychiatric 0.82 0.35d
Table 1 (Continued )
Design: Groups Effect size (g)
Reference Delusional Control Decision Certainty Pot. Disc. Reversal

f
Peters & Garety (2006) Mixed Mixed without delusions 1.49 /0.60 0.23
Peters et al. (2005)f,g Mixed Psychiatric
Effect sizes (g) are reported by study design and dependent variable. The studies included are marked in the reference section with an asterisk.
Pot.Disc. /Response to potentially disconfirmatory evidence.
a
Mortimer et al. (1996) employed a correlational design to examine the relationship between draws to decision and delusional items of the CASH.
b
Psychiatric controls had OCD.
c
The delusional group had predominately a diagnosis of schizophrenia (58%). The remainder had delusional disorder.
d
The ‘‘graded estimates’’ procedure was used. Effect sizes are for ‘‘definitely the bag with...’’ judgements, unless insufficient data was available in which
case ‘‘almost certainly the bag with . . .’’ was used (Young & Bentall, 1997, Exp. 2).
e
39% of the psychiatric control group had OCD.
f
A symptom-based approach was employed in these studies. The delusional group had mixed diagnoses, containing at least 50% patients with
schizophrenia.
g
Insufficient data were reported to compute an effect size. However, this study was included in tests in which p values were meta-analytically pooled.
JTC BIAS AND DELUSIONS

57
58 FINE ET AL.
Results and discussion

Figure 1 illustrates how published effect sizes varied by measure of JTC and
by type of nondelusional control group. Focused comparisons confirmed
that the effect size for ‘‘draws to decision’’ was significantly larger than the
effect size for ‘‘draws to certainty’’ (Z /3.79), ‘‘response to potentially
disconfirmatory evidence’’ (Z /3.14), and ‘‘response to reversal’’ (Z /4.22,
all p valuesB/.001). No significant differences were observed between any
other JTC measures (all p values/.1).4
A significant effect of comparison group was also found, with signifi-
cantly smaller effect sizes for comparisons between delusional patients and
psychiatric controls, than with nonpsychiatric controls (Z /2.16, p /.015).
To explore further the effect of comparison group, for each JTC measure the
data from the more rigorous delusional patients vs. psychiatric control
comparison were meta-analytically pooled. This revealed a significant effect
for both ‘‘draws to decision’’ (Z /6.63, p B/.001) and ‘‘draws to certainty’’
(Z /3.17, p B/.001). This latter result contrasts with the qualitative review
for this dependent variable which revealed little evidence in support of an
effect (see also Garety & Freeman, 1999). It was noted, however, that the
data from the delusional patients vs. psychiatric control comparison in the
Fear and Healy (1997) study made a substantial contribution to this effect
size (g /1.72). As noted previously, in this study the psychiatric control
group comprised patients with OCD who required significantly more ‘‘draws
to certainty’’ than did nonpsychiatric controls. Excluding the Fear and
Healy data from the analysis eliminates the significant effect for the ‘‘draws
Control Group
1.50 psychiatric
non psychiatric
Effect size, g
1.00
0.50
0.00
decision certainty potential reversal

disconfirm
Figure 1. The distribution of effect sizes (g) by type of control group and dependent variable.
4
This pattern of significance and nonsignificance also remained after Bonferroni correction for
multiple comparisons (n /6).
to certainty’’ measure (Z /1.35, p/.088) by reducing the mean effect size,

g, from 0.39 to 0.17.
There was a marginal effect for ‘‘response to potentially disconfirmatory
evidence’’ (Z /1.59, p /.056), which did reach significance for a comparison
with nonpsychiatric controls (Z /3.31, p B/.001). There were no significant
effects for ‘‘response to reversal’’, either in comparison with psychiatric
controls (Z /0.31, p/.38) or nonpsychiatric controls (Z /1.25, p/.105).5
The aim of Study 1 was to determine which measure of JTC bias best
discriminates between individuals with and without delusions. Consistent
with the qualitative review above, and with the conclusions of the Garety
and Freeman (1999) review, the effect size for the dependent variable ‘‘draws
to decision’’ was significantly greater than those for the other three putative
measures of JTC. Thus, a tendency to request fewer beads before making a
decision appears to best characterise the behaviour of people with delusions
as against those without. In line with previous conclusions (Dudley et al.,
1997b; Garety & Freeman, 1999), our findings concerning ‘‘draws to
certainty’’ suggest that delusional patients do not differ from controls in
their certainty judgements, at least as measured by the Beads task.
It is not clear how best to interpret the trend to significance observed for
the JTC measure ‘‘response to potentially disconfirmatory evidence’’. This
result provides marginal support for the idea that patients with delusions
show greater reductions in certainty than controls in response to evidence
that potentially disconfirms their hypotheses. However, it clearly does not
discriminate between the presence and absence of delusion as well as the
‘‘draws to decision’’ measure and disappears when corrected for multiple
comparisons. A decision to interpret the findings as significant, however,
would raise the question of why what appears to be a conceptually similar
measure of JTC *‘‘response to reversal’’ *does not predict the presence of
delusions. If delusional patients are thought to show greater reductions in
certainty in response to potentially disconfirmatory evidence, there is no
clear theoretical reason why they should not also change their minds more
quickly than comparison groups, in response to actually disconfirmatory
evidence. Taken together, then, the findings from the measures of ‘‘response
to potentially disconfirmatory evidence’’ and ‘‘response to reversal’’ suggest
that delusional patients are no more likely than comparison groups to ‘‘jump
to new conclusions’’.
The critical feature of patients with delusion thus appears to be specific to
a tendency to make a decision on the basis of less evidence than comparison
groups.
5
Bonferroni correction for repeated significance testing (n /8) did not affect this pattern of
results other than to confirm the nonsignificance of ‘‘response to potentially disconfirmatory
evidence’’ for comparisons with psychiatric controls.
60 FINE ET AL.
IS THE JTC BIAS AN EPIPHENOMENAL EFFECT?

The results of Study 1 suggest that the presence of psychiatric symptoma-
tology contributes to the JTC reasoning style. It further shows that the
measure ‘‘draws to decision’’ unambiguously taps an effect that supersedes
this contribution. However, a potential confound remains. The majority of
JTC studies involve patients with delusions who are diagnosed with
schizophrenia and as Menon et al. (2006), for example, have noted, the
cognitive skills commonly impaired in schizophrenia are likely to affect the
performance of reasoning tasks (e.g., memory, attention, and executive
functions; see Sharma & Harvey, 2000). Thus, a bias in reasoning observed
in delusional patients with schizophrenia may be a behavioural outcome of
one or more cognitive deficits in schizophrenia and not causal in the
formation of their delusions. In line with this epiphenomenon view, some
studies have found evidence suggestive of an interaction between schizo-
phrenic symptomatology and task demands, which may contribute to the
observed JTC effect in patients with schizophrenia (e.g., Menon et al., 2006;
Moritz & Woodward, 2005; van Dael et al., 2006).
Three approaches have been utilised to address the problem of confound-
ing factors in schizophrenia. First, and most directly, three recent studies
have either used nondelusional patients with schizophrenia as a psychiatric
control group for delusional patients (Menon et al., 2006; Moritz &
Woodward, 2005), or have used the presence of delusions to distinguish
between groups and then compared the performance of those with and
without a diagnosis of schizophrenia (Peters et al., 2005). Second, a small
number of studies have investigated reasoning in patients with delusions who
have a diagnosis of delusional disorder (Conway et al., 2002; Fear & Healy,
1997; Garety et al., 1991). The advantage of using this population is the
relative absence of possible confounding cognitive deficits. As Manschreck
(2000, p. 1254) puts it: ‘‘In general, patients with delusional disorder show
little disorganisation or impairment in their behaviour or in the clarity of
their thinking.’’ Third, some researchers have explored JTC style in
nonclinical participants who experience high levels of delusional ideation
or who are at above average risk of psychosis (Broome et al., 2003; Colbert &
Peters, 2002; van Dael et al., 2006). We review the results of each of these
approaches below.
JTC bias in schizophrenic patients with and without

delusions
The first study to include delusional and nondelusional patients with
schizophrenia was a correlational study (Mortimer et al., 1996). Mortimer
et al. gave the Beads task to a group of patients with schizophrenia whose
delusion scores ranged from 0 to 22 on delusional items of the Compre-
hensive Assessment of Symptoms and History (CASH; Andreasen, Flaum,
& Arndt, 1992). This scale assesses range of delusional beliefs as well as
degree of conviction, preoccupation, and extent to which the belief is acted
upon. Despite the wide range of delusional symptomatology reported in
their sample (from 0 to 22 on the CASH), Mortimer et al. found that 42% of
patients required only one draw to decision, a proportion similar to that
observed by Garety and colleagues. Moreover, number of draws to decision
did not correlate with the CASH delusion score. As Mortimer et al. (1996;
p. 301) noted, ‘‘[t]his suggests that abnormal probabilistic reasoning is a
consequence of having schizophrenia rather than having the particular
schizophrenic symptom of delusions’’. By contrast, however, van Dael et al.
(2006) explored the association between JTC (indexed by ‘‘extreme
responding’’) and the presence of delusions, and found evidence for a
significant link between the two in their patient group.
Menon et al. (2006) compared the performance of delusional and
nondelusional patients with schizophrenia on variants of the Beads task.
In their first experiment they found no differences on ‘‘draws to decision’’
between delusional and nondelusional patients with schizophrenia nor
indeed between any groups. There were slight procedural differences between
the Menon et al. task and that used by Huq et al. (1988) and Garety et al.
(1991). For example, in the original studies the beads, once drawn, were then
hidden from view. However, in the Menon et al. study the beads remained in
sight, thus providing a memory aid. In follow-up experiments, Menon et al.
found significant differences between the groups with schizophrenia and
nonpsychiatric controls when the original procedure was used. However,
they found no significant differences between the delusional and nondelu-
sional patients with schizophrenia. Furthermore, when the beads were left in
sight (the memory aid condition), the differences between the groups
disappeared. Menon et al. (2006; p. 533) thus suggested that ‘‘a key
component of the ‘jumping to conclusions’ (JTC) effect relates to the
memory demands imposed by the task’’. This study therefore supports the
epiphenomenon view and, in particular, identifies memory demands as a
possible contributory factor to the JTC effect observed in patients with
schizophrenia (but see Dudley et al., 1997b, who found that the addition of a
memory aid had no effect on the performance of patients with delusions).
A second recent study conducted by Moritz and Woodward (2005) also
compared the performance of delusional and nondelusional patients with
schizophrenia on the Beads task. The presence of delusional symptomatol-
ogy did not distinguish between patients with schizophrenia on ‘‘draws to
decision’’. Although Moritz and Woodward did find that delusional patients
with schizophrenia were significantly more likely than nondelusional
62 FINE ET AL.
patients to make a decision on the basis of one bead (extreme responding),

there was no significant difference between these two groups when decisions
based on the second bead were compared.
The final study to examine the issue of whether or not a JTC bias is
specific to the presence of delusions is that of Peters et al. (2005). Using a
symptom-based approach, they found that the delusional group requested
significantly fewer ‘‘draws to decision’’ in the first task (Condition 1). In
contrast to the epiphenomenon view, there was no difference between
delusional patients with and without a diagnosis of schizophrenia on ‘‘draws
to decision’’.
JTC in delusional disorder patients

The first study including delusional disorder patients conducted by Garety
et al. (1991) was somewhat inconclusive with regard to the relative
performances of delusional patients with schizophrenia and those with
delusional disorder. On the one hand, there were no significant differences in
‘‘draws to decision’’ or ‘‘responses to potentially disconfirmatory evidence’’
between these two groups (see Garety & Hemsley, 1994). This is consistent
with a specific relationship between a JTC style and the presence of
delusions, regardless of diagnosis. On the other hand, while a significant
difference was seen between the patients with schizophrenia and nonpsy-
chiatric controls on ‘‘draws to decision’’, no differences were seen between
the delusional disorder patients and either control group. This is more
consistent with the view that the JTC style is an epiphenomenon of
schizophrenic symptomatology.
Fear and Healy (1997) found that patients with delusional disorder made
significantly fewer ‘‘draws to decision’’ than an OCD group, a group with
delusions and OCD, as well as nonpsychiatric controls. However, as noted
earlier, patients with OCD may have a relatively overcautious reasoning
style, which makes the psychiatric comparison difficult to interpret. In the
third study of this sort, Conway et al. (2002) observed a reduced number of
‘‘draws to decision’’ in delusional disorder patients relative to nonpsychiatric
controls. Unfortunately, the study did not include a psychiatric control
group, leaving open the possibility that the performance of the patients with
delusions was not due to the presence of delusion but to other factors
present in psychiatric populations.
JTC bias in nonclinical populations

Research using nonclinical populations has found mixed evidence for the
role of a JTC bias in delusion. Colbert and Peters (2002) found a JTC bias in
psychiatrically healthy individuals scoring highly on the Peters et al.

Delusions Inventory (PDI; Peters, Joseph, & Garety, 1999). Since this
nonclinical population does not suffer from psychiatric symptomatology,
Colbert and Peters have argued that this suggests that the reasoning bias
observed in clinical populations is related to delusional ideation rather than
schizophrenic symptomatology. It is important to note though that there is
as yet no known relationship between scores on the PDI and the
development of clinical delusions. Therefore, while suggestive, this work
does not provide strong support for an association between a JTC bias and
the presence of clinical delusions.
Interestingly, Broome et al. (2003, personal communication) recently
found evidence of a JTC bias on hard versions of the Beads task (60:40 and
44:28:28 ratios) in individuals identified as having an ‘‘at risk mental state’’
of developing a psychotic illness (Yung et al., 1998). Since approximately
40% of young people so classified will go on to develop a psychotic illness
(e.g., Phillips, Yung, & McGorry, 2000), this provides some support for an
association between a JTC bias and vulnerability to mental illness likely to
include delusional symptomatology. Similarly, van Dael et al. (2006) recently
measured JTC (extreme responding) in four groups at increasing risk of
psychosis liability: patients with schizophrenia or schizoaffective disorder;
first degree nonpsychotic relatives of people with psychosis; and nonclinical
participants with either an above average, or average, level of psychotic
experiences. They found a dose-response relationship in the association
between psychosis liability and JTC bias. However, unlike Broome et al.
(2003) they found no evidence of a JTC bias in their ‘‘at-risk’’ nonclinical
groups (nonpsychotic first-degree relatives of patients with psychosis, and
psychiatrically healthy individuals reporting above average psychotic experi-
ences), compared with not-at-risk controls.
STUDY 2: DETERMINING THE CONTRIBUTION OF

SCHIZOPHRENIC SYMPTOMATOLOGY TO THE JTC BIAS
Given the equivocal data concerning the role of schizophrenic symptoma-
tology in delusion, we attempted to quantify its contribution to the JTC bias
using meta-analytic techniques. We tested predictions arising from both a
strong epiphenomenon view (that the JTC effect is purely a consequence of
schizophrenic symptomatology unrelated to the presence of delusions per se)
and a weak epiphenomenon view (that schizophrenic symptomatology makes
a significant contribution to the JTC effect). ‘‘Draws to decision’’ was used
to test the predictions, as this was shown in the first study to be the most
reliable measure of JTC bias.
64 FINE ET AL.

The strong epiphenomenon hypothesis. The strong epiphenomenon
hypothesis was evaluated in two ways. First, we examined evidence for a
JTC effect for schizophrenic patients with delusions in comparison with
nondelusional schizophrenic controls. Evidence for a JTC effect would
challenge the strong epiphenomenon hypothesis. When data for four studies
that evaluated the association between delusions and ‘‘draws to decision’’
within a schizophrenic population were meta-analytically pooled (Menon
et al., 2006; Moritz & Woodward, 2005; Mortimer et al., 1996; Peters et al.,
20056), evidence for a significant JTC effect was found (Z / 1.73, p/.042),
contrary to the strong epiphenomenon position.
It was not possible to test the strong epiphenomenon prediction that there
would be no significant JTC effect for delusional disorder patients in
comparison with psychiatric controls. This was because only one study
(Garety et al., 1991) used an appropriate psychiatric control group. The
second test, therefore, was for a JTC effect in nonclinical populations prone
to delusional ideation or at risk for development of delusions. ‘‘Draws to
decision’’ effect sizes were retrieved from two studies employing nonclinical
comparisons (Broome et al., 2003: g /0.90; Colbert & Peters, 2002: g/
0.80). When these data were meta-analytically pooled, evidence for a
significant JTC effect was found (Z /3.89, p/.003), contrary to the strong
epiphenomenon position. On the basis of these two results, we were able to
reject the null hypothesis that the JTC effect is purely an epiphenomenal
effect and went on to test predictions from the weak epiphenomenon
position.
The weak epiphenomenon hypothesis. We tested four predictions arising

from the weak epiphenomenon view. First, we tested the prediction that the
effect size for the comparison between nondelusional and delusional patients
with schizophrenia would be significantly smaller than the effect size arising
from the comparison between delusional patients with schizophrenia and
psychiatric controls. Figure 2 illustrates how effect size for published studies
examining JTC in delusional patients with schizophrenia varies as a function
of nondelusional control. The effect sizes for studies employing a psychiatric
control group illustrated in Figure 1 have thus been partitioned into those
involving nondelusional patients with schizophrenia and those involving
6
This study employed a symptom-based approach involving clinical groups including a high
proportion of patients with schizophrenia.
1.50
1.25
Effect Size, g
1.00
0.75
0.50
0.25
0.00
non delusional psychiatric non psychiatric

Sz
Control group
Figure 2. The distribution of ‘‘draws to decision’’ effect sizes (g) for studies comparing a group with
schizophrenia and delusions with three types of control group. Data for the nondelusional
schizophrenia control group also includes one study that employed a correlational design (Mortimer
et al., 1996).
other psychiatric groups. The meta-analysis provided support for this

prediction (Z /2.79, p /.003).7
Second, we predicted that the effect size for delusional disorder patients
compared with psychiatric controls would be significantly smaller than that
for delusional patients with schizophrenia compared with psychiatric
controls. Figure 3 illustrates how effect sizes for published studies examining
JTC in delusional disorder compared with those examining JTC in
schizophrenia. Marginal support was found for the prediction (Z /1.59,
p/.056).
Third, we tested the prediction that the effect size for delusional disorder
patients compared with nonpsychiatric controls would be significantly
smaller than that for delusional patients with schizophrenia compared
with nonpsychiatric controls. As illustrated in Figure 3, the meta-analysis
revealed no significant differences (Z /0.23, p/.41) between these effect
sizes.
Finally, we tested the prediction that the effect size for nonclinical
populations would be significantly smaller than that for delusional patients
with schizophrenia, in comparison with both psychiatric and nonpsychiatric
controls. Meta-analytic comparisons failed to support this prediction. No
significant differences were observed between effect sizes for nonclinical
comparisons and comparisons between delusional patients with schizophre-
nia and psychiatric (Z /0.23, p /.41) or nonpsychiatric controls (Z /0.28,
p/.39).
7
Peters et al. (2005) was not included in this comparison because insufficient data was reported
in this study to compute an effect size.
66 FINE ET AL.
1.50 Clinical Group

Schizophrenia
Delusional
1.25 Disorder
Effect Size, g
1.00
0.75
0.50
0.25
psychiatric non psychiatric

controls controls
Figure 3. The distribution of ‘‘draws to decision’’ effect sizes (g) by type of clinical group with
delusions and type of control group.
Our findings support the conclusion that a JTC bias, as represented by

‘‘draws to decision’’, is not entirely an artefact of schizophrenic symptoma-
tology (contrary to the strong epiphenomenon hypothesis). However, it
remains unclear whether or not schizophrenic symptomatology contributes
to a JTC style (the weak epiphenomenon hypothesis). Two of our four tests
of the weak epiphenomenon hypothesis provided support for that hypoth-
esis. When the best possible psychiatric control group is used (schizophrenic
patients without delusions), or delusional disorder patients are the target
group, the effect size for the JTC bias is more modest than that seen when
delusional patients with schizophrenia and nonschizophrenic psychiatric
controls are compared. On the other hand, the other two comparisons fail to
support the weak epiphenomenon hypothesis. The effect size for delusional
disorder patients compared with nonpsychiatric controls was no different to
that for delusional patients with schizophrenia compared with nonpsychia-
tric controls. Furthermore, the effect sizes for nonclinical populations did
not differ to that for delusional patients with schizophrenia, in comparison
with both psychiatric and nonpsychiatric controls.
Even if we set aside this latter finding, on the grounds that we do not as
yet have a clear understanding about delusional ideation in nonclinical
populations, evidence for the weak epiphenomenon hypothesis remains
equivocal, and further research will be required to resolve the discrepancy in
our findings.
We conclude this section by noting a possibility concerning the
epiphenomenon hypothesis, which we are not in a position to evaluate.
While our findings support the view that the JTC bias is not an
epiphenomenon of schizophrenia, we have not ruled out the possibility
that it is an epiphenomenon of a symptom, or a small family of symptoms,
of schizophrenia. Because patients with schizophrenia can exhibit a variety
of symptoms, and we do not have detailed symptom profiles of the patients
included in the JTC studies, the JTC bias could be an epiphenomenon of one
or more symptoms of schizophrenia overrepresented in the sample of

delusional patients studied. In order to rule out this hypothesis, it would be
necessary to consider the relation between a JTC bias and individual
symptoms of schizophrenia rather than schizophrenia as a whole (e.g.,
Garety et al., 2005; van Dael et al., 2006). While this is an onerous task,
future studies of the JTC should, where possible, make use of a more fine-
grained taxonomy of experimental subjects.
DOES EMOTIONAL SALIENCE EXACERBATE THE JTC

EFFECT?
The results of the first two studies show that ‘‘draws to decision’’ is a reliable
measure of JTC style, and that it cannot be attributed solely to psychiatric
symptomatology or to an epiphenomenal effect of the many symptoms
associated with schizophrenia. People with delusions have a tendency to
require less evidence before making a probabilistic decision.
Two concerns have been expressed regarding the JTC bias. First, it is
unclear how substantial a contribution a relatively subtle reasoning bias
could make to the production of the dramatic phenomena of delusion.
Second, the JTC bias is a reasoning style that appears to be independent of
subject matter, and this is at odds with the relatively small number of known
delusional themes (e.g., Yager & Gitlin, 2000). Regarding the first concern,
Young and Bentall (1995, p. 366) comment that, ‘‘[t]he extent to which
deficits of the magnitude observed [in their own study], or indeed the
comparable findings observed by Huq, Garety, and Hemsley (1988) or
Garety et al. (1991), could directly contribute to the formation and
maintenance of delusional beliefs, therefore remains a topic for further
investigation’’. Underlining this point is the observation that, on the neutral
Beads task, patients with delusions are actually reaching the correct
hypothesis more quickly than are controls, and that by certain standards
their reasoning could be regarded as more effective (see Maher & Spitzer,
1993). One suggestion that attempts to address these concerns is that the
‘‘jumping to conclusions’’ style in patients with delusions is exaggerated
when reasoning about material with a particular content (e.g., Dudley et al.,
1997a; Dudley & Over, 2003). For example, Dudley et al. (1997a, p. 582)
suggest that:
the effect of the pre-existing hastiness bias, in combination with additional

hastiness for material that is emotional or self-related in content, leads to more
errors being made in reasoning. Errors are therefore more likely to be made when
reasoning with emotional material. Hence it seems likely that more false beliefs may
be concerned with emotionally salient information.
68 FINE ET AL.
If emotional or self-related content enhanced a JTC bias, that might, in

one stroke, explain both how a small bias could lead to extreme beliefs, and
why delusions seem to focus on a small number of themes. Whether or not
the JTC bias is disproportionately increased in delusional patients when
forming beliefs about emotionally salient material is therefore a pertinent
question for accounts of how the JTC bias contributes to delusion
formation. It was the aim of Study 3 to address this question.
STUDY 3: DETERMINING THE EFFECT OF EMOTIONAL

SALIENCE ON THE JTC BIAS
Three studies have examined the effect of emotional salience on the JTC
bias. Dudley et al. (1997a) compared ‘‘draws to decision’’ in two conditions,
realistic-neutral and emotionally salient. All groups were hastier in the
emotionally salient condition. Further, when a realistic version of the Beads
task was used (involving students and schools or positive and negative
comments in a survey), content (neutral vs. emotionally salient) had an equal
effect on all groups. Using a similar design, Menon et al. (2006) found no
evidence that emotionally salient material increased hastiness in any of their
groups, compared with the neutral beads version of the task. In a third study
that used a ‘‘graded estimates’’ procedure, Young and Bentall (1997) found
marginal support for a Group / Salience interaction, but this reflected a
trend for a greater effect of emotional salience in both the delusional and
psychiatric control groups, compared with the nonpsychiatric group.
In order to clarify the putative effect of emotional salience on the JTC
bias, we conducted a meta-analysis.

Table 2 summarises the 14 effect sizes yielded from the three studies
investigating the role of emotional saliency on variants of the Beads task.
Contrary to the expectations of the ‘‘emotional saliency’’ position, effect
sizes for the comparison between patients with delusions and nonpsychiatric
controls were not significantly greater for emotionally salient stimuli when
compared with realistic-neutral (Z /0.33, p /.37), neutral beads (Z / 0.22,
p/.41), or neutral stimuli combined (realistic neutral plus neutral beads;
Z /1.07, p /.14). Similarly, no significant differences in effect sizes for
comparisons with psychiatric controls were found between emotionally
salient stimuli and realistic-neutral (Z /0.63, p /.26), neutral beads (Z /
0.24, p/.41) or neutral stimuli combined (Z /0.75, p/.23).
These results thus reveal no statistical support for the idea that a
‘‘jumping to conclusions’’ reasoning style is disproportionately exacerbated
TABLE 2
Summary of effect sizes yielded from three published studies investigating the role
of emotional saliency on variants of the Beads task
Effect size (g)
Reference Control group Neutral Realistic neutral Emotionally salient
Dudley et al. (1997a)a Psychiatric 0.76 1.00

Young & Bentall (1997, Exp. 2)b Psychiatric 0.08 0.05
Menon et al. (2006, Exp. 1)a Psychiatricc 0.05 0.02 0.24
Nonpsychiatric 0.03 /0.27 /0.22
The delusional group in each case were patients with schizophrenia.

a
Dependent variable was ‘‘draws to decision’’.
b
Composite effect sizes were computed as the arithmetic mean of ‘‘draws to certainty’’, ‘‘draws
to reversal’’, and ‘‘response to potentially disconfirmatory evidence’’.
c
The psychiatric control group were nondelusional patients with schizophrenia.
by emotional material in patients with delusions. In line with this conclusion

is the recent finding that a similar percentage of patients with delusions were
‘‘extreme responders’’ (required only one or two beads) on the neutral Beads
task as were ‘‘extreme responders’’ on an emotionally salient version of the
task (Garety et al., 2005).
GENERAL DISCUSSION
Three conclusions are supported by the current findings. First, ‘‘draws to
decision’’ is the most reliable measure of JTC bias. This offers some
important constraints on our theoretical understanding of the JTC reason-
ing style. Second, the JTC bias, as represented by ‘‘draws to decision’’,
appears to make a genuine contribution to delusional symptomatology, and
is not solely an epiphenomenal effect. This provides evidence for the
possibility of a causal role for a JTC reasoning style in the formation and/
or retention of delusional belief. Third, emotional salience does not increase
the hastiness of decision-making in patients with delusions. This finding
suggests that the particular delusional beliefs adopted by a patient cannot be
explained by the emotional nature of those types of belief. Rather, delusional
patients may be hasty with respect to all types of material.
In the following sections we consider two questions about the JTC bias.
First, what are the implications of our findings for the underlying basis of
the JTC bias? And, second, what role does the JTC bias play in the
70 FINE ET AL.
development of delusion? With respect to the latter question, we consider a

possible role both in the formation and the retention of delusion.
Implications for the underlying basis of the JTC bias

In line with the conclusions of previous authors (e.g., Dudley et al., 1997b;
Garety & Freeman, 1999), the present results support the view that a critical
idiosyncrasy in reasoning amongst patients with delusions is a willingness to
make a decision on the basis of less evidence than nondelusional groups;
however, delusional patients do not differ with respect to certainty of
judgement. Our findings also suggest that in response to conflicting
evidence, delusional patients are not especially inclined to ‘‘jump to new
conclusions’’; nor are they especially recalcitrant with respect to conclusions
they have already reached.
Only two of the prominent accounts of the cognitive disturbance
underlying the JTC bias are consistent with our current findings. Both
predict that delusion will be associated only with reduced ‘‘draws to
decision’’. The first is Bentall and colleagues’ suggestion that the JTC bias
arises from a difficulty in processing sequential information (e.g., Young &
Bentall, 1995). The second is Dudley and Over’s (2003) proposal that
delusional patients have a confirmatory reasoning style.
In contrast, accounts based on the propensity of individuals with
schizophrenia to be overinfluenced by current stimuli at the expense of
prior learning (e.g., Kapur, 2003; Gray, Feldon, Rawlins, Hemsley, & Smith,
1991) have difficulty explaining why patients with delusions do not ‘‘jump to
new conclusions’’. Menon et al.’s (2006) variant of this proposal, which
suggests an inappropriate weighting of evidence, is inconsistent with the
normal certainty ratings seen in delusional groups. Further, our findings do
not fit with the idea that delusional patients have a high ‘‘need for closure’’,
as conceptualised by Kruglanski and Webster (1996), since we found no
strong evidence that they are more certain about their judgements, or that
they cling tenaciously to those judgements in the face of disconfirmatory
evidence.
A possible role for the JTC bias in the development of

delusion
Delusion formation. As indicated above, our results appear to restrict the
current theories of the JTC mechanism either to that of Young and Bentall’s
(1997) hypothesis that the JTC bias derives from a difficulty in the
processing of sequential information, or to that of Dudley and Over
(2003), according to which delusional patients have a confirmatory reason-
ing style. If either of these hypotheses is true, then the JTC bias would appear
to come into play after a delusional hypothesis has already been formulated.
Sequential information does not seem to play any role in the development of
a hypothesis but only in the evaluation of evidence for or against it. Thus, on
Bentall and colleagues’ account, a delusional hypothesis must already be
present before a JTC bias would be manifested in delusional cognition.
Similarly, a confirmatory reasoning style can make an appearance only after
a hypothesis has been formulated and is available to be confirmed or
rejected. Thus, on Dudley and Over’s account as well, a JTC bias could only
play a role in delusion development once a delusional hypothesis occurs to
the patient.
In particular, on Dudley and Over’s (2003) account, while people without
delusions apply a confirmatory reasoning style to danger-related conditional
statements, delusional patients apply such a style even to neutral situations
because they perceive danger and threat in them. It is a normal reasoning
strategy applied to situations in which only the delusional individual
erroneously perceives threat. Thus the patient must presumably first develop
a delusional sense of threat or danger prior to extending their confirmatory
reasoning style to neutral situations. This position is supported by our
finding that the JTC effect is not disproportionately enhanced by emotion-
ally salient material. Thus it cannot be that a JTC bias, applied to an
emotionally salient domain, can ‘‘boot-strap’’ the development of a
delusional belief. The situation being reasoned about must first gain a
delusional significance before the JTC bias can get a purchase.
A delusional thought must therefore be present before a JTC bias can
have an effect on the development of delusion. A JTC bias may, however,
lead to a premature acceptance of such a thought. Evidence for this view
comes from Garety et al. (2005) who have recently found evidence that a
JTC style of reasoning is associated with delusional conviction. Their
analysis supported a model in which ‘‘belief inflexibility’’ *a reluctance to
consider the possibility that one might be wrong about one’s delusional
belief*largely mediates the contribution of a JTC reasoning style to
delusional conviction. The role of JTC in this account appears to be one
of facilitating the precipitous acceptance of the delusional hypothesis, and
‘‘precluding reflection on past learning to consider whether the information
fits with previous knowledge, with the result that the possibility that one
might be mistaken is not considered’’ (p. 381). In line with this position,
Freeman et al. (2004) found that patients who did not report any alternative
explanations for the experiences on which their delusions were based were
more likely to jump to conclusions than were patients who could come up
with alternative explanations.
There is another reason for limiting the role of the JTC bias in this way,
and this is the strikingly small number of forms of delusion that exist. Yager
and Gitlin (2000, p. 801) list only 14 ‘‘classic’’ forms, and over 85% of
72 FINE ET AL.
delusions observed in patients with schizophrenia fall into one of four

delusional forms (Gutiérrez-Lobos, Schmid-Siegal, Bankier, & Walter, 2001;
Jørgensen & Jensen, 1994). Identical delusional forms recur across the large
number of psychotic and nonpsychotic disease states in which delusions
occur (Jørgensen & Jensen, 1994). Moreover, the existence of only a small
number of delusional forms is not only uniform across disease states, but
also across cultures (e.g., Gutiérrez-Lobos et al., 2001; Kim et al., 2001;
Stompe et al., 1999). A domain-general JTC bias, such as acts even on the
neutral Beads task, cannot easily explain this feature of delusional beliefs.
For example, there is presumably any number of bizarre beliefs an individual
with a ‘‘jumping to conclusions’’ information processing style could jump to.
Even if the claim that reasoning disorders are exacerbated by emotionally
salient material were supported by the data, one would still expect to observe
a very large number of delusion forms. Thus it is not obvious how the
domain-general JTC bias could be involved in the genesis of the limited
number of actual forms. A better explanation of this feature of delusion may
come from the proposal that in order to make sense of unusual internal or
external events we turn to one of a small number of explanatory frameworks
(e.g., Zimbardo, 1999).
Delusion maintenance. We have found that patients with delusions do

not seem to treat evidence that potentially disconfirms their hypothesis
differently to nondelusional controls. This appears to be in conflict with
Moritz and Woodward’s (2005) suggestion that patients with schizophrenia
have a specific impairment in the processing of disconfirmatory evidence *
the ‘‘bias against disconfirmatory evidence’’ (BADE). Our findings thus
leave us with no obvious explanation of the tenacity of delusional beliefs,
except that all beliefs that are personally significant are held tenaciously.8
There is no evidence at the moment, therefore, that the JTC bias plays any
role in the maintenance of a delusional belief once it has been accepted.
Overview
Taken together, our findings suggest that a causal role for a JTC bias should
be hypothesised to lie in that stage of delusion development in which a
patient is considering whether or not to accept or reject a thought with a
delusional content. The JTC bias does not seem to play any role in either the
production of a delusional thought or in its maintenance once it has been
accepted. We conclude that further research on the processes of delusion
acceptance may illuminate the role of a JTC bias in delusion.
8
We are grateful to one of our reviewers for this point.
We note that three issues could fruitfully be addressed in the investigation

of delusional belief acceptance. The first concerns insight in delusion. There
is evidence that a significant proportion of patients with delusion are able to
recognise deficiencies in the reasoning associated with their own delusional
beliefs. One study (Jørgensen, 1995) found that some two-thirds of
individuals with delusions exhibit insight from the onset of their illness.
On the face of it, to have insight into one’s delusion is to recognise that the
belief one takes to be plausible enough to be embraced should not be
embraced. In the context of a JTC bias, this seems to mean that one can
recognise that the evidence that one has taken to be sufficient to adopt a
delusion is, in fact, insufficient. In order to have insight, then, it seems that
one must have reasoning abilities that are unbiased, and this is at odds with
the bias that is implicated in the acceptance of the delusion to begin with.
One possible solution to this difficulty would be to establish that insight
amounted to a recognition that one’s delusion was odd or bizarre rather
than a recognition that the evidence supporting the delusional belief was
insufficient. The JTC approach to explaining delusion would be enhanced by
investigating whether insight can coexist with a JTC bias, and by a deeper
understanding of the exact nature of that insight.
A second topic for investigation concerns the way in which hypotheses
come to be entertained. A hypothesis must first be deemed worthy of
investigation before confirmatory or disconfirmatory evidence will be
sought. It is a common experience among nondelusional individuals to
have a thought that is immediately rejected as a nonstarter and for which
evidence is never sought. Persecutory thoughts, for example, occur to
nondelusional individuals but are regularly rejected as implausible. The
question arises, therefore, why delusional patients take delusional
hypotheses *particularly those that are, on the face of it, extremely
implausible *as serious contenders for acceptance. Since the JTC bias
concerns the relation of evidence to the delusional hypothesis, it is unlikely
to be relevant to this issue. As Freeman and colleagues have hypothesised, an
individual’s emotional state and pre-existing beliefs about herself, others and
the world are likely to contribute to the content of her explanations of
anomalous experiences, and whether particular explanations continue to be
entertained (Freeman, Garety, Kuipers, Fowler & Bebbington, 2002).
A third issue worthy of investigation is the nature of evidence in delusion.
For a JTC bias to play a role in delusion acceptance, the patient must be able
to identify putative evidence on the basis of which to jump to a delusional
conclusion. The identification of evidence is likely to involve an abnormal
cognitive process distinct from the JTC bias itself because there can be no
good evidence for some delusions. For example, were a nondelusional
individual to entertain the hypothesis that a thought had been inserted into
his mind, it is unlikely that any putative evidence would be taken to be good
74 FINE ET AL.
evidence. In contrast, if a JTC bias plays a role in delusion formation, there

must be evidence that the patient takes to be good evidence on the basis of
which they can jump to the delusional conclusion. A complete account of
delusional belief acceptance, therefore, must be able to explain how
delusional individuals identify the evidence for a JTC bias to work on.
Since the JTC bias concerns the quantity of evidence required for the
acceptance of a delusion, it is unlikely to be relevant to an understanding of
how evidence comes to be chosen by the delusional individual in the first
place.
Conclusions
The results of our meta-analyses provide support for the view that patients
with delusions exhibit a genuine difference in the amount of evidence they
require to embrace a hypothesis, and that this difference is not entirely a
consequence of the presence of schizophrenia. The JTC bias does not extend
to the processing of contradictory information following the formation of a
hypothesis, and it is not disproportionately increased when the content of
material being reasoned about is emotionally salient.
Overall, consideration of our findings and the possible role the JTC bias
could take in delusion, suggests that the JTC bias is not relevant to the
formulation of delusional hypotheses. In particular, it does not appear to be
relevant to the question of why delusional forms are so small in number. Nor
is the JTC bias likely to play a role either in an explanation of why patients
take these implausible hypotheses seriously or in the explanation of how
delusional patients identify putative evidence for the hypotheses. Finally, the
JTC bias seems to have no role to play in the maintenance of delusion in the
face of disconfirmatory evidence. Further investigation into the processes of
belief acceptance *both pathological and nonpathological *may prove
fruitful in illuminating the way in which the JTC bias exerts its effect in
delusional patients.
Manuscript received 12 November 2003
Revised manuscript received 29 March 2006
REFERENCES
(Studies included in the meta-analysis are preceded by an asterisk.)
American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th
ed.). Washington, DC: Author.
Andreasen, N. C., Flaum, M., & Arndt, S. (1992). The Comprehensive Assessment of Symptoms
and History (CASH): An instrument for assessing diagnosis and psychopathology. Archives of
General Psychiatry, 49 , 615 623.
Bentall, R. P., Corcoran, R., Howard, R., Blackwood, N., & Kinderman, P. (2001). Persecutory
delusions: A review and theoretical integration. Clinical Psychology Review, 21 , 1143 1192.
Brankovic, S. B., & Paunovic, V. R. (1999). Reasoning under uncertainty in deluded schizophrenic
patients: A longitudinal study. European Psychiatry, 14 , 76 83.
*Broome, M., Brett, C., Johns, L., Woolley, J., Peters, E., Garety, P., & McGuire, P. K. (2003).
Reasoning styles and delusions in early psychosis. Schizophrenia Research , 60 (Suppl. 1), 12.
*Colbert, S. M., & Peters, E. R. (2002). Need for closure and jumping-to-conclusions in delusion-
prone individuals. Journal of Nervous and Mental Disease, 190 , 27 31.
*Conway, C. R., Bollini, A. M., Graham, B. G., Keefe, R. S., Schiffman, S. S., & McEvoy, J. P.
(2002). Sensory acuity and reasoning in delusional disorder. Comprehensive Psychiatry, 43 ,
175 178.
*Dudley, R. E., John, C. H., Young, A. W., & Over, D. E. (1997a). The effect of self-referent
material on the reasoning of people with delusions. British Journal of Clinical Psychology, 36 ,
575 584.
*Dudley, R. E., John, C. H., Young, A. W., & Over, D. E. (1997b). Normal and abnormal
reasoning in people with delusions. British Journal of Clinical Psychology, 36 , 243 258.
Dudley, R. E. J., & Over, D. E. (2003). People with delusions jump to conclusions: A theoretical
account of research findings on the reasoning of people with delusions. Clinical Psychology and
Psychotherapy, 10 , 263 274.
*Fear, C. F., & Healy, D. (1997). Probabilistic reasoning in obsessive-compulsive and delusional
disorders. Psychological Medicine, 27 , 199 208.
Freeman, D., Garety, P. A., Fowler, D., Kuipers, E., Bebbington, P. E., & Dunn, G. (2004). Why do
people with delusions fail to choose more realistic explanations for their experiences? An
empirical investigation. Journal of Consulting and Clinical Psychology, 72 , 671 680.
Freeman, D., Garety, P. A., Kuipers, E., Fowler, D., & Bebbington, P. E. (2002). A cognitive model
of persecutory delusions. British Journal of Clinical Psychology, 41 , 331 347.
Garety, P., Kuipers, E., Fowler, D., Freeman, D., & Bebbington, P. (2001). A cognitive model of the
positive symptoms of psychosis. Psychological Medicine, 31 , 189 195.
Garety, P. A., & Freeman, D. (1999). Cognitive approaches to delusions: A critical review of
theories and evidence. British Journal of Clinical Psychology, 38 , 113 154.
Garety, P. A., Freeman, D., Jolley, S., Dunn, G., Bebbington, P. E., Fowler, D. G., et al. (2005).
Reasoning, emotions, and delusional conviction in psychosis. Journal of Abnormal Psychology,
114 , 373 384.
Garety, P. A., & Hemsley, D. R. (1994). Delusions: Investigations into the psychology of delusional
reasoning . Oxford, UK: Oxford University Press.
*Garety, P. A., Hemsley, D. R., & Wessely, S. (1991). Reasoning in deluded schizophrenic and
paranoid patients: Biases in performance on a probabilistic inference task. Journal of Nervous
and Mental Disease, 179 , 194 201.
Gray, J. A., Feldon, J., Rawlins, J. N. P., Hemsley, D. R., & Smith, A. D. (1991). The
neuropsychology of schizophrenia. Behavioral and Brain Sciences, 14 , 1 84.
Gutiérrez-Lobos, K., Schmid-Siegel, B., Bankier, B., & Walter, H. (2001). Delusions in first-
admitted patients: Gender, themes and diagnoses. Psychopathology, 34 , 1 7.
Hedges, L. V. (1981). Distribution theory for Glass’s estimator of effect size and related estimators.
Journal of Educational Statistics, 6 , 107 128.
*Huq, S. F., Garety, P. A., & Hemsley, D. R. (1988). Probabilistic judgements in deluded and non-
deluded subjects. Quarterly Journal of Experimental Psychology, 40A , 801 812.
Jørgensen, P. (1995). Recovery and insight in schizophrenia. Acta Psychiatrica Scandinavica , 92 ,
436 440.
Jørgensen, P., & Jensen, J. (1994). Delusional beliefs in first admitters: A clinical description.
Psychopathology, 27 , 100 112.
76 FINE ET AL.
Kapur, S. (2003). Psychosis as a state of aberrant salience: A framework linking biology,

phenomenology, and pharmacology in schizophrenia. American Journal of Psychiatry, 160 ,
13 23.
Kim, K., Hwu, H., Zhang, L. D., Lu, M. K., Park, K. K., Hwang, T. J., et al. (2001). Schizophrenic
delusions in Seoul, Shanghai and Taipei: A transcultural study. Journal of Korean Medical
Science, 16 , 88 94.
Kruglanski, A., & Webster, D. (1996). Motivated closing of the mind: ‘‘Seizing’’ and ‘‘freezing’’.
Psychological Review, 103 , 263 283.
Linney, Y. M., Peters, E. R., & Ayton, P. (1998). Reasoning biases in delusion-prone individuals.
British Journal of Clinical Psychology, 37 , 285 302.
Maher, B. A., & Spitzer, M. (1993). Delusions. In P. B. Sutker & H. E. Adams (Eds.),
Comprehensive handbook of psychopathology (pp. 263 293). New York & London: Plenum
Press.
Manschreck, T. C. (2000). Delusional disorder and shared psychotic disorder. In B. J. Sadock & V.
A. Sadock (Eds.), Comprehensive textbook of psychiatry (7th ed., Vol. 1, pp. 1243 1264) .
Philadelphia: Lippincott, Williams, & Wilkins.
McGuire, L., Junginger, J., Adams, S. G., Jr., Burright, R., & Donovick, P. (2001). Delusions and
delusional reasoning. Journal of Abnormal Psychology, 110 , 259 266.
*Menon, M., Pomarol-Clotet, E., McKenna, P. J., & McCarthy, R. A. (2006). Probabilistic
reasoning in schizophrenia: A comparison of the performance of deluded and nondeluded
schizophrenic patients and exploration of possible cognitive underpinnings. Cognitive
Neuropsychiatry, 11 , 521 536.
Menon, M., Woodward, T. S., Pomarol-Clotet, E., McKenna, P. J., & McCarthy, R. A. (2005).
Heightened stimulus salience renders schizophrenics less susceptible to the ‘‘famous names
illusion’’. Schizophrenia Research , 80 , 369 371.
*Moritz, S., & Woodward, T. S. (2005). Jumping to conclusions in delusional and non-delusional
schizophrenic patients. British Journal of Clinical Psychology, 44 , 193 207.
Moritz, S., Woodward, T. S., & Hausmann, D. (2006). Incautious reasoning as a pathogenetic
factor for the development of psychotic symptoms in schizophrenia. Schizophrenia Bulletin , 32 ,
327 331.
*Mortimer, A., Bentham, P., McKay, A., Quemada, I., Clare, L., & Eastwood, N. (1996).
Delusions in schizophrenia: A phenomenological and psychological exploration. Cognitive
Neuropsychiatry, 1 , 289 303.
Mosteller, F. M., & Bush, R. R. (1954). Selected quantitative techniques. In G. Lindzey (Ed.),
Handbook of social psychology: Vol. 1. Theory and method . Cambridge, MA: Addison-Wesley.
*Peters, E., & Garety, P. (2006). Cognitive functioning in delusions: A longitudinal analysis.
Behaviour Research and Therapy, 44 , 481 514.
Peters, E. R., Joseph, S. A., & Garety, P. A. (1999). Measurement of delusional ideation in the
normal population: Introducing the PDI (Peters et al. Delusions Inventory). Schizophrenia
Bulletin , 25 , 553 576.
*Peters, E. R., Thornton, P., Siksou, L., Linney, Y., & MacCabe, J. (2005). Specificity and origins of
the ‘‘jump-to-conclusions’’ bias in deluded patients. Manuscript submitted for publication .
Pélissier, M. C., & O’Connor, K. P. (2002). Deductive and inductive reasoning in obsessive-
compulsive disorder. British Journal of Clinical Psychology, 41 , 15 27.
Phillips, L., Yung, A., & McGorry, P. (2000). Identification of young people at risk of psychosis:
Validation of Personal Assessment and Crisis Evaluation Clinic intake criteria. Australian and
New Zealand Journal of Psychiatry, 34 (Suppl.), 164 169.
Rector, N. A., & Beck, A. T. (2002). Cognitive therapy for schizophrenia: From conceptualization
to intervention. Canadian Journal of Psychiatry, 47 , 39 48.
Rosenthal, R. (1991). Meta-analytic procedures for social research . London: Sage.
Rosenthal, R., & Rubin, D. B. (1982). Comparing effect sizes of independent studies. Psychological
Bulletin , 92 , 500 504.
Sharma, T., & Harvey, P. D. (Eds.). (2000). Cognition in schizophrenia: Impairments, importance,
and treatment strategies. Oxford, UK: Oxford University Press.
Stompe, T., Friedman, A., Ortwein, G., Strobl, R., Chaudhry, H. R., Najam, N., & Chaudhry, M.
R. (1999). Comparison of delusions among schizophrenics in Austria and in Pakistan.
Psychopathology, 32 , 225 234.
Tracz, S. M., Elmore, P. B., & Pohlmann, J. T. (1992). Correlational meta-analysis: Independent
and non-independent cases. Educational and Psychological Measurement , 52 , 879 888.
Van Dael, F., Vermissen, D., Janssen, I., Myin-Germeys, I., van Os, J., & Krabbendam, L. (2006).
Data gathering: Biased in psychosis? Schizophrenia Bulletin , 32 , 341 351.
Volans, P. J. (1976). Styles of decision making and probability appraisal in selected obsessional and
phobic patients. British Journal of Social and Clinical Psychology, 15 , 305 317.
Yager, J., & Gitlin, M. J. (2000). Clinical manifestations of psychiatric disorders. In B. J. Sadock &
V. A. Sadock (Eds.), Comprehensive textbook of psychiatry (7th ed., Vol. 1, pp. 789 823) .
Philadelphia: Lippincott, Williams, & Wilkins.
Young, A. W. (2000). Wondrous strange: The neuropsychology of abnormal beliefs. Mind and
Language, 15 , 47 73.
Young, H. F., & Bentall, R. P. (1995). Hypothesis testing in patients with persecutory delusions:
Comparison with depressed and normal subjects. British Journal of Clinical Psychology, 34 ,
353 369.
*Young, H. F., & Bentall, R. P. (1997). Probabilistic reasoning in deluded, depressed and normal
subjects: Effects of task difficulty and meaningful versus non-meaningful material. Psycholo-
gical Medicine, 27 , 455 465.
Yung, A., Phillips, L., McGorry, P., McFarlane, C., Francey, S., & Harrigan, S. (1998). Prediction
of psychosis: A step towards indicated prevention of schizophrenia. British Journal of
Psychiatry, 172 (Suppl. 33), 14 20.
Zimbardo, P. G. (1999). Discontinuity theory: Cognitive and social searches for rationality and
normality *may lead to madness. Advances in Experimental Social Psychology, 31 , 345 486.

FDFHFDH

Uploaded by

Copyright:

Available Formats

FDFHFDH

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FDFHFDH

Uploaded by

Copyright:

Available Formats

COGNITIVE NEUROPSYCHIATRY

2007, 12 (1), 46 77

Hopping, skipping or jumping to conclusions? Clarifying

Introduction . There is substantial evidence that patients with delusions exhibit a

INTRODUCTION: THE ‘‘JUMPING TO CONCLUSIONS’’ BIAS

WHAT IS THE JTC BIAS?

evidence’’ and ‘‘response to reversal’’. After all, one feature of delusional

Theoretical accounts of the JTC bias

patients will have greater confidence in their hypotheses, or that their

STUDY 1: A META-ANALYTIC COMPARISON OF

found that delusional patients showed significantly smaller ‘‘responses to

recent articles in this area, were used to identify potentially relevant

Design: Groups Effect size (g)

Reference Delusional Control Decision Certainty Pot. Disc. Reversal

Huq et al. (1988) Schizophrenia Psychiatric 0.94

Design: Groups Effect size (g)

Reference Delusional Control Decision Certainty Pot. Disc. Reversal

JTC BIAS AND DELUSIONS

Results and discussion

decision certainty potential reversal

to certainty’’ measure (Z /1.35, p/.088) by reducing the mean effect size,

IS THE JTC BIAS AN EPIPHENOMENAL EFFECT?

JTC bias in schizophrenic patients with and without

patients to make a decision on the basis of one bead (extreme responding),

JTC in delusional disorder patients

JTC bias in nonclinical populations

psychiatrically healthy individuals scoring highly on the Peters et al.

STUDY 2: DETERMINING THE CONTRIBUTION OF

Results and discussion

The weak epiphenomenon hypothesis. We tested four predictions arising

non delusional psychiatric non psychiatric

other psychiatric groups. The meta-analysis provided support for this

1.50 Clinical Group

psychiatric non psychiatric

Our findings support the conclusion that a JTC bias, as represented by

or more symptoms of schizophrenia overrepresented in the sample of

DOES EMOTIONAL SALIENCE EXACERBATE THE JTC

the effect of the pre-existing hastiness bias, in combination with additional

If emotional or self-related content enhanced a JTC bias, that might, in

STUDY 3: DETERMINING THE EFFECT OF EMOTIONAL

Results and discussion

Effect size (g)

Reference Control group Neutral Realistic neutral Emotionally salient

Dudley et al. (1997a)a Psychiatric 0.76 1.00

The delusional group in each case were patients with schizophrenia.

by emotional material in patients with delusions. In line with this conclusion

development of delusion? With respect to the latter question, we consider a

Implications for the underlying basis of the JTC bias

A possible role for the JTC bias in the development of

delusions observed in patients with schizophrenia fall into one of four

Delusion maintenance. We have found that patients with delusions do

We note that three issues could fruitfully be addressed in the investigation

evidence. In contrast, if a JTC bias plays a role in delusion formation, there

Kapur, S. (2003). Psychosis as a state of aberrant salience: A framework linking biology,

You might also like