Paranoia Research Paper

This is a repository copy of A systematic review of self-report measures of paranoia.
White Rose Research Online URL for this paper:

http://eprints.whiterose.ac.uk/133248/
Version: Accepted Version
Article:
Statham, V., Emerson, L.-M. and Rowse, G. orcid.org/0000-0003-3292-4008 (2019) A
systematic review of self-report measures of paranoia. Psychological Assessment, 31 (2).
pp. 139-158. ISSN 1040-3590
https://doi.org/10.1037/pas0000645
© 2019 American Psychological Association. This paper is not the copy of record and may
not exactly replicate the authoritative document published in the APA journal. Please do
not copy or cite without author's permission. The final article is available, upon publication,
at: 10.1037/pas0000645
Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless
indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by
national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of
the full text version. This is indicated by the licence information on the White Rose Research Online record
for the item.
Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request.
eprints@whiterose.ac.uk
https://eprints.whiterose.ac.uk/
Abstract
Paranoia can be conceptualised as consisting of a hierarchy of cognitions, ranging from
commonly experienced thoughts about less severe perceived threats, up to less common,
persecutory thoughts about extreme threats, which are associated with distressing psychosis.
This review systematically appraises self-report paranoia questionnaires validated for use
among the general population; the type of paranoia assessed, measurement or psychometric
properties, and subsequent validation with clinical samples are all considered. A systematic
literature search was performed using PubMed, Web of Science, and PsycInfo databases.
Study methodologies and measurement properties were evaluated according to COnsenus-
based Standards for the selection of health-based Measurement Instruments (Mokkink et al.,
2012). Twenty-six studies, describing the validation of nine paranoia-related questionnaires,
were identified. Questionnaires were reviewed in relation to the hierarchy of paranoia; with
two questionnaires assessing ‘low-level’ paranoia, four assessing persecutory thoughts, and
the remainder assessing paranoia across this continua. Questionnaires assessing the full
hierarchy of paranoid thoughts, alongside associated dimensions such as pre-occupation,
conviction, and distress, offer the most comprehensive assessment of paranoia in both non-
clinical and clinical populations. Of the measures which do this, the Green et al. (2008)
Paranoid Thoughts Scale had the strongest evidence for its measurement properties and is
therefore recommended as the most reliable and valid self-report assessment of paranoia
currently available. However, this review illustrated that generally paranoia questionnaires
lack high quality evidence for their measurement properties. Implications of these findings
for clinical practice and research are discussed.
Keywords: paranoia; self-report; questionnaires; validation; Psychometric; properties;
measures; assessment
Public significance statement
1
This systematic review identified nine self-report questionnaires that have been developed to
assess paranoia and were designed for use with the general population. An analysis of studies
that used these questionnaires suggested that the Green et al. (2008) Paranoid Thoughts Scale
has the best evidence for the reliability and validity of its test scores.
Introduction
Paranoia has traditionally been conceptualised as a symptom of psychosis-related
diagnoses such as ‘Schizophrenia’ (American Psychiatric Association, 2013). However,
increasing evidence shows that as with other clinically-relevant experiences (e.g. obsessive-
intrusive thoughts; Berry & Laskey, 2012; voice hearing; Beavan, Read, & Cartwright,
2011), paranoid thoughts are also experienced by those without mental health difficulties.
Freeman (2006) reviewed studies assessing different types of paranoid thoughts, in general
population samples, over different time periods, and found varying estimates of the
prevalence of paranoid cognitions, ranging from approximately 2% to 42%. Data from a
large, nationally-representative UK sample has also found paranoid thinking to be associated
with a variety of difficulties, including poorer physical and mental health (e.g. anxiety, worry,
insomnia, suicidal ideation), reduced social functioning, lack of social support, and increased
use of alcohol and cannabis (Freeman et al., 2011). While there may be debate as to whether
these associated difficulties contribute towards the development of paranoia, or are a
consequence of it, they highlight the potential gains of the study of this phenomenon, both to
individuals affected by paranoid experiences, and to wider society.
What constitutes a ‘paranoid’ experience is often not well defined within the
literature. Bentall et al. (2009) suggest that paranoia occurs as a result of a combination of
cognitive biases (e.g. threat anticipation, jumping to conclusions, difficulty understanding
others’ mental states), and accompanying “emotion-related processes, such as anxiety,
depression, and self-esteem” (p. 244). Paranoia can also be conceptualised as a more stable
2
personality trait (often described as “suspiciousness”), that can vary between individuals and
within theories of schizotypy is associated with an increased vulnerability to develop
psychotic symptoms (Johns & van Os, 2001). Accordingly, examples of extreme paranoid
personality traits are associated with diagnoses such as ‘Paranoid Personality Disorder
(American Psychiatric Association, 2013). While paranoia may be associated with certain
“suspicious” personality presentations, this review focuses in more detail upon the specific
types of threat-based cognitions, thoughts and beliefs that could be described as ‘paranoid’.
Understanding the ideational experience of paranoia in isolation, rather than as part of a more
stable personality structure, accounts for the fact that paranoid thoughts can fluctuate from
moment to moment (Ben-Zeev, Ellington, Swendsen, & Granholm, 2010; Nittel et al., 2018)
and can decrease in response to psychological interventions (Freeman & Garety, 2014).
What makes a cognition ‘paranoid’ is often not well defined, and thoughts can be
described as ‘paranoid’ and ‘persecutory’ synonymously, when the meaning of these
descriptors may differ (McKay, Langdon, & Coltheart, 2006). Academic research often
focuses upon ‘persecutory’ thoughts, which are defined as explicit concerns about threats of
current/ongoing harm to oneself, enacted by an intentional perpetrator (Freeman & Garety,
2000). However, paranoia can be conceptualised more broadly as including thoughts relating
to an exaggerated and “persistent misconception of one-self as the target of another’s
thoughts or action” (Fenigstein & Vanable, 1992, p. 130), often described as ‘ideas of
reference’.
Persecutory thoughts and paranoid ideas of reference are distinguished somewhat
from concerns about threats to society or wider social groups (e.g. broader conspiracy
theories) by the focus of threat to oneself. However, there is likely to be overlap between
these constructs. Ideas of reference more broadly could also be part of non-paranoid
psychological difficulties, such as the self-focused attention seen among socially anxious
3
individuals (Clark & Wells, 1995), or those with “grandiose delusions” (e.g. relating to
inflated sense of worth, or a special identity, Knowles, McCarthy-Jones, & Rowse, 2011).
This perhaps explains why paranoia and ideas of reference emerged as distinct facets of
positive schizotypy in Cicero and Kern’s (2010) factor analytic study. What distinguishes
paranoid ideas of reference from other self-referential thoughts may be how these thoughts
are appraised, and whether they are associated with assumptions of ill will, hostility, or
suspicious intent (Fenigstein & Vanable, 1992). Accordingly, questionnaires that only assess
ideas of reference include items that may have positive appraisals and thus would not
necessarily assess paranoia (e.g. thinking that people are waving at you - The Referential
Thinking Scale, Lenzenweger, Bennet, & Lilenfield, 1996), as well as those more related to
paranoia, which are likely to imply hostility,. In support of the importance of self-referential
thoughts within the construct of paranoia, Stefanis et al. (2004) also demonstrated that ideas
of reference load on to a paranoia factor, along with social anxiety and suspiciousness.
Freeman et al.’s (2005) hierarchy of paranoia provides a framework to organise these
different paranoid cognitions, and includes thoughts that are less explicitly persecutory (e.g.
ideas of reference), within a broad conceptualisation of paranoia. Freeman et al. (2005) order
their identified paranoid cognitions in a hierarchy according to the severity of perceived
threat, beginning with social evaluative concerns at the bottom (defined as interpersonal
worries, such as fears of rejection/vulnerability/the world being dangerous), followed by
ideas of reference, and finally persecutory thoughts relating to mild (e.g. people trying to
cause irritation), moderate (e.g. people going out of their way to get at you), then severe (e.g.
people trying to cause you significant harm) threats, at the top of the hierarchy. There are
other theoretical models of paranoid cognition, which generally seek to explain persecutory
beliefs in isolation and focus upon the origin and maintenance of these experiences (e.g.
Bentall et al., 2009; Freeman, Garety, Kuipers, Fowler, & Bebbington, 2002). Furthermore,
4
as previously discussed, there are schizotypy theories which describe ‘suspicious’ personality
traits (see Grant, Green, & Mason, 2018 for review) that may be associated with a greater
incidence of paranoid cognitions. Thus, the paranoia hierarchy is the only widely cited model
known to the authors that provides a structured account of the types of thought content can be
said to be part of paranoid experience, and it is therefore used to structure this review.
Subsequent studies examining paranoid thoughts from Freeman et al.’s (2005)
hierarchy have found that those from the lower hierarchy emerge as a distinct factor from
persecutory thoughts (from the upper hierarchy) within factor analyses (Green et al., 2008;
Ibáñez-Casas et al., 2015). Ideas of reference and social evaluative concerns are proposed as
being the building blocks for the development of more explicit persecutory thoughts, and thus
assessing both types of cognition alongside each other is argued to provide a more
comprehensive understanding of paranoid ideation (Freeman et al., 2005). Relatedly, many
studies find strong associations between paranoia and self-focused or self-conscious cognitive
styles (Combs & Penn, 2004; Freeman et al., 2012; Smári, Stefánsson, & Thorgilsson, 1994).
Green et al. (2008) and Ibáñez-Casas et al. (2015) found that ideas of reference in
social situations (social reference thoughts) were the most commonly endorsed paranoid
thoughts among the general population, whereas persecutory ideas were the most commonly
endorsed paranoid thoughts among clinical participants. Nevertheless, both types of thoughts
were much more prevalent among individuals with persecutory delusions (PDs), suggesting
that the entire hierarchy has clinical relevance to paranoia (Green et al., 2008).
There may be also factors other than paranoid thought content that influence whether
these cognitions are clinically-relevant experiences. For example, paranoid thoughts that are
more frequent, distressing, and appraised with more conviction and preoccupation are more
common among clinical populations (Green et al., 2008; Ibáñez-Casas et al., 2015). Indeed,
Peters, Joseph, Day, and Garety (2004) argue that the distress, conviction, and preoccupation
5
associated with persecutory cognitions determine how ‘delusion-like’ they are. Alternatively,
Trower and Chadwick (1995) distinguish ‘poor me’ paranoia, where persecution is perceived
as unjust or undeserved, and ‘bad me’ paranoia, where persecution is perceived as a deserved
consequence of an individual’s actions. Research suggests that ‘poor me’ paranoia is more
common among those with psychosis-related diagnoses (Melo & Bentall, 2013; Melo,
Concoran, Shryane, & Bentall, 2009).
Much of our understanding of paranoia among both clinical and non-clinical
populations has been obtained using self-report questionnaires. Within these questionnaires
paranoia is defined and assessed differently, which is may have influenced endorsement rates,
and contributed towards the varying prevalence estimates for delusions and paranoid
cognitions in the general population (Freeman, 2006). As persecutory thoughts are more
common among clinical samples, and ideas of reference are more common among non-
clinical samples as opposed to those experiencing psychosis (Green et al., 2008; Ibáñez-Casas
et al., 2015), prevalence estimates are likely to be influenced by both the type of paranoid
thought content from the hierarchy of perceived threat (Freeman et al., 2005) that is being
assessed, and the population to which the questionnaire is administered.
Aside from within large symptom inventories (e.g. Minnesota Multiphasic Personality
Inventory-2 Restructured Form; Ben-Porath & Tellegen, 2011), there are no paranoia specific
self-report measures developed primarily with clinical samples. Rather, diagnostic interview
tools tend to be preferred (e.g. Composite Diagnostic Interview; Kessler, Andrews, Mroczek,
Ustun, & Wittchen, 1998). This preference stems from the historical use of diagnosis to
classify distressing psychotic experiences and also arguments (summarised by Bell, Fiszdon,
Richardson, Lysaker,& Bryson, 2007) that those experiencing psychosis may struggle to self-
report accurately due to holding unusual beliefs, experiencing cognitive deficits, or desiring
to minimise their experiences (e.g. due to stigma, as a defensive coping strategy). The
6
evidence for these arguments is mixed, and varies based upon the construct assessed
(Baumstark et al., 2013; Kim et al., 2010; Selton, Wiersma, & van den Bosch, 2000). For
PDs, Lincoln, Ziegler, Lüllmann, Müller, and Rief (2010) found a strong relationship
between self-reported and observed-rated experiences, whereas Liraud, Droulout, Parrot, and
Verdoux (2004) did not. However, a lack of association between self-reported and observer-
rated paranoia does not necessarily indicate that an observers are more accurate than those
self-reporting their experiences. Furthermore, self-report assessments have additional
advantages such as their ability to be distributed widely, with fewer resources required, and
potentially less impact of social desirability bias compared with a face-to-face interview.
The primary aim of this review is to critically evaluate existing self-report measures
that were developed to assess paranoia with general population samples. However, it is also
acknowledged that the inclusion of individuals with psychosis in the development and
subsequent validation of these measures is important, to evidence the construct validity of
scales and examine whether items are clinically-relevant. Indeed, not evidencing the clinical
relevance of items has been a criticism when assessing other constructs in the general
population, such as obsessive-intrusive thoughts (Berry & Laskey, 2012). Furthermore,
questionnaires validated clinically and non-clinically have greater potential utility. Psychotic-
like experiences that occur without significant distress or impairment increase the later risk of
symptoms that may warrant a clinical diagnosis (Hanssen, Bak, Bijl, Vollebergh, & van Os,
2005; Welham et al., 2009). Thus, assessing paranoia across clinical and non-clinical
populations could highlight variables that increase the likelihood of paranoia-related distress.
Questionnaires validated for use with individuals experiencing distressing paranoia could also
be used clinically within assessment or for outcome measurement within interventions.
In this review, we therefore aim to provide a critical appraisal of the measurement
properties of self-report measures of paranoia that were developed using non-clinical
7
participants, or a mixed clinical and non-clinical group. Additionally, studies validating
questionnaires with clinical populations, that were originally developed with non-clinical or
mixed samples were included. The inclusion of these measures therefore encompasses the
full continua of experience through non-clinical to clinical As well as evaluating the
measurement properties of these questionnaires, we aim to use Freeman et al.’s (2005)
proposed hierarchy of perceived threat, as a framework to categorise the construct of paranoia
that is assessed by each measure. Thus, the reviewed questionnaires relate to
conceptualisations of ‘low level’ paranoia, from the lower hierarchy (e.g. Paranoia Scale,
Fenigstein & Vanable, 1992), ‘persecutory beliefs’, from the upper hierarchy (e.g.
Persecutory Ideation Questionnaire, McKay et al., 2006) and paranoia constructs that span
the entirety of the hierarchy (e.g. Green et al. Paranoid Thoughts Scale, Green et al., 2008).
Method
As this paper describes a literature review, no ethical approval was required for the
research.
Search Strategy
The methods undertaken in this review were informed by guidelines in the Preferred
Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement 2009
(Moher, Liberati, Tetzlaff, & Altman, 2009). A systematic search using PubMed, Web of
Science, and PsycInfo databases was performed on January 4, 2017 (see Appendix A,
supplementary materials). Synonyms of terms for the construct of interest (e.g. paranoia,
suspiciousness, persecutory), population for questionnaire development (e.g. general
population, non-clinical, community), instrument type (e.g. questionnaire, scale, inventory),
and questionnaire properties (e.g. psychometric, reliability, validity), were used to search the
titles, abstracts, and keywords of publications. Papers containing keywords for comorbid
difficulties associated with paranoia (e.g. dementia, Parkinson’s) were excluded.
8
Initial searching identified 2432 papers. Firstly duplicate papers were removed from
the search results (n = 707), followed by articles which after abstract and title screening did
not meet the inclusion criteria (n = 1667). The full text of remaining papers (n = 58) was
screened, followed by an ancestry search of studies included after this stage. Database and
ancestry searching was used to find papers pertaining to both the original development and
subsequent psychometric validation of the identified questionnaires. A citation search was
also performed for studies documenting the initial development of each measure. The
screening and data extraction process was completed by the primary author.
Inclusion and Exclusion Criteria
The following inclusion criteria were applied: 1. Studies must describe a self-report
questionnaire rather than observer/interview-based assessments; 2. Studies must describe the
initial development of a questionnaire, or indicate within the abstract that the aim is to
validate the measurement properties of the questionnaire; 3. Studies must assess measurement
properties outlined by Terwee, de Vet, Prinsen, and Mokkink (2011), or complete item-
response theory (IRT) analyses (Kean & Reilly, 2014), or latent class analyses (Dayton &
Macready, 2006); 4. Included questionnaires must have a scale or subscale for the assessment
of paranoia or persecutory delusions. Furthermore, all questionnaire subscales must assess
constructs relating to paranoia or delusions. Questionnaires measuring a range of delusions
must have a specific persecutory delusion subscale, and present psychometric data relating to
this subscale specifically. Thus, in-keeping with the review’s specific focus upon the
assessment of paranoid ideation, paranoia subscales were not included if they were part of
larger questionnaires either measuring non-delusional elements of psychosis (e.g. voice
hearing, negative symptoms), other mental health difficulties (e.g. depression, personality
disorder), or personality traits; 5. In line with Freeman and Garety (2000), the paranoia
assessed must relate to fears of present/ongoing harm to the self (rather than to society, or
9
social groups); 6. Questionnaires must have been developed using at least a small proportion
of participants who were recruited from the general population; 7. Studies evaluating the
measurement properties of existing questionnaires (i.e. not the original development papers),
may include samples from any population (e.g. clinical/non-clinical/mixed); 8. Studies must
describe questionnaires developed to assess paranoia among adults (aged 18 +). However,
articles describing questionnaires originally developed with adult populations, then applied to
younger samples (aged 14 +), were included; 9. Articles must be published in peer-review
journals; 10. Full articles must be available in English.
Articles were excluded if: 1. The questionnaires solely measured cognitive biases
involved in paranoia, or reactions to/appraisals of paranoid experiences; 2. Questionnaires
assessed paranoia solely in relation to another condition or difficulty - thus not assessing
paranoia distinctly, but its overlap with other constructs. For example, dementia, Parkinson’s
disease, substance misuse, learning disability, paranoid personality disorder, or depression.
Quality Appraisal
The COnsenus-based Standards for the selection of health-based Measurement
Instruments (COSMIN) protocol for systematic reviews of self-report questionnaires (Terwee
et al., 2011) was followed to appraise the measurement properties of questionnaires.
COSMIN definitions of measurement properties were therefore adopted, as were COSMIN
standards for how to appraise these properties. The COSMIN appraisal tool was developed by
systematically reviewing existing criteria for good measurement properties, following which
a multi-disciplinary panel of experts reached a consensus upon which properties to include
within the tool, and how their quality would be judged (Mokkink, Terwee, Patrick, et al.,
2010). Thus, the COSMIN protocol was deemed a comprehensive, systematically developed
framework, that was grounded in the knowledge of experts. Mokkink, Terwee, Gibbons et al.
(2010) evaluated the inter-rater reliability of COSMIN appraisal ratings and found 80%
10
agreement between raters on at least two thirds of items. Adjustments to the tool and manual
were made to address areas where reliability was weaker (Mokkink et al., 2012). The
COSMIN tool has now been applied within numerous systematic reviews aiming to appraise
the quality of questionnaires (e.g. Sutton et al., 2016; Wigham & McConachie, 2014).
To establish the quality of papers included in this review the quality of the
methodologies used to assess measurement properties was firstly appraised. The COSMIN
tool (Mokkink et al., 2012) appraises methodologies which assess different forms of
reliability, namely internal consistency, test-retest reliability, and measurement error; as well
as different types of validity, namely content/face validity, criterion validity, and construct
validity (includes structural validity, testing of hypotheses about related/unrelated constructs,
and cross-cultural validity). The responsiveness of measures and also IRT methodologies can
also be evaluated. The appraisal items used to assess each measurement property are provided
in the supplementary materials (Appendix B). However, items included how missing items
were handled, study samples sizes, and whether the unidimensionality of scales was
evidenced (e.g. for internal consistency). For each applicable appraisal item, studies were
rated ‘poor’, ‘fair’, ‘good’, or ‘excellent’. Following the recommended ‘worst score counts’
procedure, the lowest item rating was taken to represent the overall methodological quality of
analyses establishing that measurement property (Terwee et al., 2012).
COSMIN definitions of measurement properties were followed. For example, while
some studies claimed to evidence criterion validity by comparing clinical and non-clinical
groups on their paranoia scores, COSMIN defines these analyses as assessing construct
validity. If the methodology used to establish a measurement property was cited within a
different paper, where possible this was obtained and consulted for the required information.
The methodology for content validity was rated if a questionnaire was being validated for the
first time, or with a new population (e.g. a new culture or clinical population).
11
Once the methodological quality of psychometric analyses had been appraised, the
second stage was to appraise whether the psychometric findings themselves met the
recommended standards (e.g. internal consistency: Cronbach’s alpha ≥ .7). An
accompaniment to the COSMIN methodological checklist was used, covering the same
aspects of reliability, validity, and responsiveness (Terwee et al., 2011). Each measurement
property was assessed positively, negatively, or indeterminately. The standards required for
each measurement property, are provided in supplementary materials, Appendix C.
An overall rating for the strength of each measurement property, for each
questionnaire, was created by combining the methodological quality appraisal score for a
measurement property with ratings for the quality of the psychometric property itself.
Evidence was rated as either positive (+) or negative (-), and the strength of evidence in either
direction was rated according to the categories shown in Table 1.
Table 1
Ratings for the strength of evidence for each measurement property
Level Rating Criteria
Strong +++ or --- Consistent findings in multiple studies of good
methodological quality OR in one study of excellent
methodological quality
Moderate ++ or -- Consistent findings in multiple studies of fair
methodological quality OR in one study of good
methodological quality
Limited + or - One study of fair methodological quality
Conflicting +/- Conflicting findings
Unknown ? Only studies of poor methodological quality
Indeterminate I All included studies reported indeterminate findings
Note. + = positive evidence, and - = negative evidence. Indeterminate category created by the author.
Adapted from COSMIN website: COSMIN.nl
A second, independent researcher (postgraduate trainee clinical psychologist)
conducted the quality appraisal procedure for studies (n = 7) relating to three randomly
selected paranoia questionnaires. Quality appraisal was similarly conducted by combining the
appraisal of the study methodology with an appraisal of the psychometric findings reported to
obtain an overall rating. Inter-rater reliability for the ratings of overall strength of evidence
for measurement properties was good (Kvalseth, 1989), with a Cohen’s Kappa = .80.
12
Disagreements were resolved through discussion and consultation with COSMIN
recommendations (Terwee et al., 2012). Initial ratings were then adjusted if necessary.
Results
Twenty-six papers were identified which described the measurement properties of
nine different paranoia-related questionnaires (Table 2): the Paranoia Scale (PS; Fenigstein &
Vanable, 1992), Paranoia/Suspiciousness Questionnaire (PSQ; Rawlings & Freeman, 1996),
Persecutory Ideation Questionnaire (PIQ; McKay et al., 2006), Persecution and Deservedness
Scale (PaDS; Melo et al., 2009), Peters et al. Delusions Inventory (PDI; Peters, Joseph, &
Garety, 1999), State Social Paranoia Scale (SSPS; Freeman et al., 2007), Paranoia Checklist
(PC; Freeman et al., 2005), State Paranoia Checklist (SPC; Schlier, Moritz, & Lincoln, 2016),
and the Green et al. Paranoid Thoughts Scale (GPTS; Green et al., 2008).
The remaining papers retrieved reported adapted versions of these measures or further
validated their measurement properties: PS (Barreto Carvalho et al., 2014; Combs, Penn, &
Fenigstein, 2002; Smári et al., 1994), PSQ (Huppert, Smith, & Apfeldrof, 2002), PIQ (Jones,
Fernyhough, de-Wit, & Meins, 2008; Van Dongen, Buck, Koole, & Van Marle, 2011), PDI
(Cella, Sisti, Rocchi, & Preti, 2011; Jones & Fernyhough, 2007; Jung et al., 2008; Lincoln et
al., 2010; López-Ilundain, Pérez-Nievas, Otero, & Mata, 2006; Peters et al., 2004; Prochwicz
& Gaw da, 2015; Rocchi et al., 2008; Verdoux et al., 1998), PC (Lincoln et al., 2010; Moritz,
Van Quaquebeke, & Lincoln, 2012), and GPTS (Ibáñez-Casas et al., 2015). Lincoln et al.
(2010) presented psychometric evaluations of both the PDI and the PC and findings were
considered separately for each measure.
For all nine questionnaires, the lead author reviewed the content of items and the
construct of paranoia that the authors of the measure claimed to assess. This allowed the
examination of how the themes of questionnaire items related to Freeman et al.’s (2005)
paranoia hierarchy and questionnaires were categorised based upon this model.
13
Table 2
Paranoia questionnaires identified and evidence reported for their measurement properties
Author & Year Construct of # items Sample Paranoia subscales Measurement properties
location paranoia/PDs
Paranoia Scale
Fenigstein & ‘Normal’, ‘non- 20 Four different student IC. All samples ≥ .81
Vanable 1992 pathological’ paranoia. samples, n ranged R. r = .70 (n = 180)
[1] Suspiciousness/assumpti from 119 to 180 HT. Associations with measures of trust (rs ≥ .30 ≤
United States ons of hostility .32), experience/inward expression of anger (rs ≥ .45 ≤
of America reminiscent of clinical .51), outwardly expressed anger (r = .18*), belief in
(USA) paranoia, occurring control of others (r = .34**) and need for personal
independent of control (r = .29**)
psychiatric problems SV. 1-factor structure explaining 25% of the variance
(N = 581)
Smári et al. 1994 20 N = 30 IC. = .87
[2] Patients with HT. Associations with a feeling of being watched (r =
Iceland schizophrenia .27**) and scores on a clinician-rated measure of
diagnoses paranoia (r = .51**)
Combs et al. 2002 20 n = 191 (non- IC. non-Hispanic Whites, = .88, African-Americans,
[3] Hispanic Whites) = .79
USA n = 102 (African- HT. The two ethnic groups differed similarly on the PS
Americans) and clinical measurements of paranoia: African-
Students American students had significantly higher levels of
paranoia (for all comparisons p < .005)
Barreto 2014 20 N = 1218 Adolescent Mistrust thoughts (8-items), IC. ≥ .72 for subscales
Carvalho et high school pupils persecutory ideas (8-items), SV. 3-factor structure explaining 46.6% of variance
al. aged 14 to 22 self-depreciation (3-items)
[4]
Portugal
Paranoia/Suspiciousness Questionnaire
Rawlings & 1996 Paranoia/suspiciousness 47 n = 264 (Sample 1) Interpersonal IC. = .87 (total scale, n = 297), ranged between
Freeman among the non- n = 297 (Sample 2) suspiciousness/hostility (12- 64 to .89 for subscales (N = 561)
[5] psychiatric population. Students items), negative mood/withdrawal SV. 5-factor structure (N = 561)
Australia (7-items), anger/impulsiveness R. r = .82 (n = 74)
(9-items), mistrust/wariness (6-
items), perceived
hardship/resentment (7-items)
Six-items had no subscale
14
Paranoia/Suspiciousness Questionnaire
Rawlings & 1996 Paranoia/suspiciousness 47 n = 264 (Sample 1) Interpersonal IC. = .87 (total scale, n = 297), ranged between
Freeman among the non- n = 297 (Sample 2) suspiciousness/hostility (12- 64 to .89 for subscales (N = 561)
[5] psychiatric population. Students items), negative mood/withdrawal SV. 5-factor structure (N = 561)
Australia (7-items), anger/impulsiveness R. r = .82 (n = 74)
(9-items), mistrust/wariness (6-
items), perceived
hardship/resentment (7-items)
Six-items had no subscale
Huppert et 2002 n = 33 (patients with IC. Total scale ≥.85 for both samples
al. schizophrenia-related R. r = .67 (n = 23)
[6] diagnoses) n = 46 HT. Positive, statistically significant (p < .05)
USA (patients with correlations with scores on 9 different self-report
anxiety/depression) measures of anxiety and depression: rs ≥ .32 ≤ .73
Persecutory Ideation Questionnaire
McKay et al. 2006 ‘Persecutory’ ideation 10 n = 98 (students) IC. = .87 (students) and .90 (patients)
[7] n = 25 (patients with HT. Positively correlated with PSQ scores of students
Australia experience of PDs) (r = .85***) and clinical participants (r = .85***)
Correlation with observer-rated PDs among clinical
participants (r = .61***). Insignificant correlation
between PSQ scores and observed-rated PDs when PIQ
scores were partialed out (r = -.14), versus significant
correlations between observed-rated PDs and PIQ
scores with PSQ scores partialed out (r = .51*)
Jones et al. 2008 Reduce n = 183 (PIQ e- IC. ≥. 84 for PIQ-7 and PIQ-10 (paper and online
[8] from 10 questionnaire) versions)
United to 7- n = 188 (paper- SV. 1-factor structure, excluding three items from
Kingdom items version of PIQ) original measure, demonstrated with both samples
(UK) Students
Van Dongen et 2011 10 n = 269 (community IC. = .78 (community sample) and .89 (clinical
al sample) sample)
[9] n = 88 (individuals R. ICC = .82 (n = 38, community participants)
Holland with schizophrenia- HT. Positively correlated with self-reported positive
related diagnoses) psychotic symptoms (r = .51***), but removing
persecutory items from the comparison measure hardly
affected this correlation (minimal divergence), r =
.51***. Significantly higher PIQ scores among clinical
15
participants (U = 256.00**)
Persecution and Deservedness Scale
Melo et al. 2009 Persecutory beliefs and 10 n = 318 (British Persecution beliefs and Analyses using combined British/Portuguese sample:
[10] the perceived students) deservedness beliefs relating to IC. = .84 (Persecution). For deservedness calculated
UK/Portugal ‘deservedness’ of n = 290 (Portuguese the same 10-items an ICC = .38.
persecution. students) SV. 1-factor structure explaining 42% of the variance
n = 45 (patients with (Persecution subscale). 1-factor structure (deservedness
PDs) subscale)
HT. Persecution scores correlated strongly with PS
scores (r s = .78***) and self-reported depression (r s =
.57***). Deservedness scores correlated moderately
with PS scores (r s = .28***) and self-reported
depression (r s = .35***). Significantly higher
persecution scores for patients as opposed to students
(p < .001).
CCV. ‘Substantially identical’ factor structures for
British and Portuguese samples independently
Peters et al. Delusions
Peters et al. 1999 PDs in the general 40 N = 272. (students 5-item subscale designed to SV. 11-factor structure explaining 59% of the variance
[11] population. Attenuated and researcher assess PDs. However, factor
UK versions of delusions acquaintances) analysis found three paranoia-
related subscales: persecution (5-
items), suspiciousness (3-items),
and paranoid ideation (4-items)
Items assessed on dimensions of
conviction, pre-occupation and
distress
Verdoux et 1998 21 N = 444 (GP surgery One PD-related subscale: SV. 7-factor structure explaining 55.3% of the variance
al. [12] attendees) “suspiciousness and persecutory
France ideas” (4-items)
Peters et al. 2004 21 N = 444 (university Two items selected from each of SV. Select the two highest loading items from each
[13] staff, students and the three PD-related, factor factor identified by Peters et al. (1999) to create a
UK research analytically identified subscales shortened questionnaire
acquaintances) by Peter’s et al. (1999)
16
Autor & Year Construct of # items Sample Paranoia subscales Measurement properties
Peters et al. Delusions Inventory
Jung et al. 2004 40 N = 310 (community Initially identify “persecutory SV. 10-factor structure explaining 57% of the variance.
[14] sample) ideas” and “jealousy and However, they argue that the dominant factor suggests
Korea suspiciousness” subscales - do not a unidimensional structure (un-rotated explains 26% of
state number of items variance)
The authors later conclude that a
unidimensional scale is more
appropriate
Jones & 2007 21 N = 493 (students) Dispute the existence of IC. Verdoux et al.’s (1998) suspiciousness and
Fernyhough previously established paranoia- persecutory ideas subscale ( = .50)
[16] related subscales López-Ilundain et al.’s (2006) paranoid subscale ( =
UK .26)
SV. Lack of “valid multifactorial structure”
López- 2006 21 N = 356 (community Factor analysis identified a SV. 7-factor structure explaining 53.7% of the variance
Illudain et sample) “paranoid” subscale (2-items)
al.
[15]
Spain
Rocchi et al. 2008 21 n = 89 (outpatients Refer to a “paranoia dimension” For combined clinical/non-clinical sample: largest class
[17] with psychosis- of the PDI (4-items) found in latent class analysis (n = 140; 41.1%) related
Italy related diagnoses to a high probability of endorsing PDI items from the
n = 210 (community paranoia dimension
sample)
Lincoln et 2010 40 N = 80 (patients with Peters et al. (1999) original 5- HT. Positively correlated with observer-rated PDs (r =
al. [18] psychosis-related item PD scale .34***)
Germany diagnoses)
Cella et al. 2011 21 n = 400 (British) For combined British/Italian sample: latent class
[19] n = 400 (Italian) analysis identified a class (n = 330; 41.3%) associated
UK & Italy Community samples with endorsement of two items with paranoid themes
17
Peters et al. Delusions Inventory
Prochwicz & 2015 40 N = 421 (community Initially identified subscales for SV. 14-factor structure explaining 58.68% of variance.
Gaw da sample) ‘suspiciousness’, ‘ideation of However, scree plot suggests a unifactorial structure
[20] persecution and body distortion’,
Poland and ‘ideation of persecution’ -
number of items not reported
The authors later argue for a
unidimensional scale
State Social Paranoia Scale
Freeman et 2007 Assesses the expectation 10 n = 100 (community IC. ≥ .84 for all samples
al. of harm from an sample) R. ICC = .74 (n = 42)
[21] intentional perpetrator in n = 64 (students) HT. Positively correlated with interviewer-rated
UK a recent situation n = 21 (those at high paranoia (r = .73***), GPTS scores (r = .41***), visual
risk of developing analogue paranoia (r = .59***) and character hostility
psychosis) ratings (r = .63***) in the community sample, and PS
scores in the student (r = .31*) and clinical (r = .44*)
samples. Negatively correlated with perceptions of VR
characters as positive (r = -.27***) or neutral (r = -
.44***)
Paranoia Checklist
Freeman et 2005 Assesses paranoid 18 N = 1202 (students) Items rated on dimensions of IC. ≥ .90 for all rating scales
al. thoughts of a “more frequency, conviction and distress HT: Positively correlated with PS frequency (r =
[22] clinical nature” than the .71***), conviction (r = .62***), and distress (r =
UK PS .58***) scores
Lincoln et al. 2010 18 N = 80 (patients with Items rated on dimensions of HT. Observer-rated PDs positively correlated with PS
[18] psychosis-related frequency, conviction and distress frequency (r = .43**), conviction (r = .39**), and
Germany diagnoses) distress (r = .38**) scores
Moritz et al. 2012 18 N = 1899 ‘Unspecified suspiciousness’ (11- SV. 2-factor structure explaining 64% of the variance
[23] (community sample) items) and ‘psychotic paranoia’
Germany (5-items)
2 items had no subscale.
18
Author & Construct of # items Sample Paranoia subscales Measurement properties
State Paranoia Checklist
Schlier et al. 2006 State-adapted version of 13, 5, n = 1893 (community Sample 1:
[24] the PC assessing paranoia and 3- sample 1) SV. 1-factor structure for all versions
Germany “in the moment”, rather item n = 1966 (community RSP. Change effect size for 13-item, d = .17, 5-item, d
than as a trait version sample 2) = .19, and 3-item SPCs, d = .27
HT: All versions of the PC were correlated with trait
measures of paranoia (rs ≥ .47 ≤ .55) and measures of
social anxiety ( rs ≥ .42 ≤ .46). Within a regression, PS
frequency and distress scores were significantly
predicted by anxiety, anger, depression and shame, but
not significantly predicted by joy.
Sample 2:
IC: ≥. 74 for all versions
Green et al. Paranoid Thoughts Scale
Green et al. 2008 Assesses a hierarchy of 32 n = 353 (university Persecution (16-items) and social FA. On pool of 95 items. 2-factor structure explaining
[25] paranoid thoughts; from staff or students) reference (16-items) Items rated 49.7% of the variance (non-clinical sample). 16-tems
UK social reference thoughts n = 50 (individuals on dimensions of preoccupation, per factor retained
to persecutory ideas. with PDs) conviction, and distress IC. ≥ .90 for both samples on both subscales
R. ICC ≥ .80 for all subscales (n = 164, non-clinical)
HT. For both samples all GPTS scales were positively
correlated with other measures of paranoia (rs ≥ .35 ≤
.86) anxiety (rs ≥ .34 ≤ .49) and depression (rs ≥ .42 ≤
.60) Significantly higher scores for clinical participants
(p < .001)
RSP. GPTS change scores correlated with change
scores on interview-based paranoia measure (n = 30,
clinical sample)
Ibáñez- 2015 32 n =151 (community Persecution (16-items) and social IC: ≥ .90 for both samples on all subscales
Casas et al. sample) reference (16-items) SV: 2-factor structure explaining 61.7% of the variance
[26] n = 40 (patients with (non-clinical sample)
Spain delusions) HT. Positively correlated with PDI (Smaller correlation
with anxiety and depression measures. Higher scores
for clinical group**: cut off of 92 gives 97.35%
specificity and 65% sensitivity.
Note. *p < .05, **p < .01, ***p < .001, IC = Internal Consistency, R = Reliability, CV = Content Validity, SV = Structural Validity, HT = Hypothesis Testing, CCV
= Cross-Cultural Validity, RSP = Responsiveness. ICC = intraclass correlation coefficient. For FA explained variance is included in the table where reported.
19
The PS (Fenigstein & Vanable, 1992) and PSQ (Rawlings & Freeman, 1996) measure
commonly occurring paranoia among the general population, rather than so-called
‘pathological’ paranoia. These measures were deemed to best assess the lower levels of
Freeman et al.’s (2005) paranoia hierarchy. Conversely, the PIQ (McKay et al., 2006), PaDS
(Melo et al., 2009), PDI (Peters, Joseph, & Garety, 1999), and SSPS (Freeman et al., 2007)
assess persecutory ideas, from the top levels of the paranoia hierarchy. The PaDS also
assesses the perceived deservedness of persecution and the PDI assesses delusion-like
qualities of persecutory ideas (conviction, pre-occupation, and distress). The PC (Freeman et
al., 2005), SPC (Schlier, Moritz, & Lincoln, 2016), and the GPTS (Green et al., 2008), assess
paranoia across the hierarchy, including ideas of reference and persecutory ideation.
Quality Analysis
Methodological quality ratings for each paper are shown in Table 3, along with
ratings illustrating the overall strength of evidence for the measurement properties of each
questionnaire. A full breakdown of the methodological ratings can be requested from the
author.
Many studies did not describe how missing data were handled. As this can be a source
of bias (Mokkink et al., 2012), such study methodologies were not rated better than ‘fair’.
Furthermore, the limited piloting of questionnaires meant that content validity and cross-
cultural validity methodologies were rated ‘poor’, and no good psychometric evidence for
these properties was reviewed. Methodologies for assessing structural validity and testing
construct validity hypotheses were relative strengths for many studies, and accordingly these
properties received stronger ratings. No studies assessed measurement error, criterion
validity, or used IRT. Only two studies assessed the responsiveness of a questionnaire to
measure change over time. Finally, no questionnaires included embedded validity indicators
to assess the accuracy of the self-reported experiences.
20
Table 3
Quality ratings for study methodologies and ratings for overall evidence for the measurement properties
Internal Reliability Content Structural Hypothesis Cross-cultural validity Responsiveness
consistency validity validity testing
Paranoia Scale Icelandic, Portuguese & African-American
samples
Evidence for measurement property ++ - + -- ++ ?
Methodological quality of studies
Fenigstein & Vanable (1992) Fair Fair Good Fair Poor
Smári et al. (1994) Poor Poor Fair Poor*
Combs et al. (2002) Poor Poor Fair Poor
Barreto Carvalho et al. (2014) Fair Poor Fair Poor*
PSQ
Evidence for measurement property - + I I ?
Rawlings & Freeman (1996) Fair Fair Fair Fair
Huppert et al. (2002) Poor Poor Poor Poor
PIQ Dutch sample
Evidence for measurement property + + ? I + ?
McKay et al. (2006) Poor Poor Fair
Jones et al. (2008) Fair Fair
Van Dongen et al. (2011) Poor Fair Poor Poor Poor
(continued)
21
PaDS Portuguese sample
Evidence for measurement property P: ++ ? -- P: + ?
D: ? D: -
Melo et al. (2009) P: Good Poor Good Fair Poor
D: Poor
PDI French, Spanish, Korean, Italian, German, and
Polish samples
Evidence for measurement property - ? +/- - ?
Peters et al. (1999) Poor Good
Verdoux et al. (1998) Poor Fair Poor
Peters et al. (2004) Poor
Jung et al. (2008) Poor Fair Poor
López-Illundain et al. (2006) Poor Fair Poor
Jones & Fernyhough (2007) Fair Fair
Rocchi et al. (2008) CNR Poor*
Lincoln et al. (2010) CNR Fair Poor*
Cella et al. (2011) CNR Poor*
Prochwicz & Gaw da (2015) Poor Fair Poor
(continued)
22
SSPS
Evidence for measurement property ? + ? +
Freeman et al. (2007) Poor Fair Poor Fair
PC German sample
Evidence for measurement property ? ? + +/- I
Freeman et al. (2005) Poor Poor Fair
Lincoln et al. (2010) CNR Fair Poor*
Moritz et al. (2012) CNR Fair Fair*
SPC German sample
Evidence for measurement property + CNR I + I ?
Schlier et al. (2016) Fair CNR Fair Fair Fair* Poor
GPTS Spanish sample
Evidence for measurement property ? ++ ++ ? ++ ? ?
Green et al. (2008) Poor Good Poor Poor Fair Poor
Ibáñez-Casas et al. (2015) Poor Excellent Poor Fair Poor
Note. +++ or --- (strong positive or negative evidence), ++ or – (moderate positive or negative evidence), ‘+ or – (limited positive or negative
evidence), +/- (conflicting findings), ? (only studies of poor methodological quality), or I (quality not possible to determine). P = persecution subscale.
D = deservedness subscale. Although all papers were written in English, for some properties information needed to rate methodological quality was
contained in another non-English language paper. In such cases either no items in an appraisal section could be rated (CNR) or ratings were based on a
subset of items (*). Cross-cultural validity was rated for studies using measures in a different language or culture. Blank cells indicate where a
measurement property was not examined within a paper.
23
24
Measures Assessing the Lower Paranoia Hierarchy
The PS (Fenigstein & Vanable, 1992) and the PSQ (Rawlings & Freeman, 1996) were
designed to assess ‘normal’, ‘non-pathological’ paranoia, and the items in the questionnaire
best reflect social evaluative concerns and ideas of reference within the lower levels of
Freeman et al.’s (2005) hierarchy. Items from both measures also go beyond the hierarchy,
assessing constructs related to paranoia, such as self-depreciation (PS; Barreto Carvalho et al.
2014) and anger/impulsiveness (PSQ; Rawlings & Freeman, 1996). Persecutory ideas from
further up the paranoia hierarchy are not assessed; described as “obviously psychotic”
(Fenigstein & Vanable, 1992, p. 131) and less relevant to the ‘normal’ population.
Rawlings and Freeman (1996) identified a 5-factor structure for the PSQ without
stating the explained variance. The factor structure of the PS was also unclear, with
Fenigstein and Vanable (1992) retaining a 1-factor structure, whereas with Portuguese
adolescents, Barreto Carvalho et al. (2014) retained a 3-factor structure. These conflicting
results could reflect methodological problems with the initial factor analysis of the measure,
or limited validity of scores across age or these cultures. Evidence for the cross-cultural
validity of the PS in Portuguese was poor, as studies did not conduct factor analyses to
replicate the structural validity of scores on the measure and used samples dissimilar to the
original development sample (Combs et al., 2002; Smári et al., 1994). The PS had evidence
of good internal consistency of its test scores and some mixed findings with regards to
construct validity (Barreto Carvalho et al., 2014; Combs et al., 2002; Fenigstein & Vanable,
1992; Smári et al., 1994). However, its test-retest reliability correlations were not adequate
(Fenigstein & Vanable, 1992). Test-retest reliability was evidenced for scores on the PSQ.
However, no other measurement properties were rated positively, which was often due to
methodological limitations (Rawlings & Freeman, 1996).

25
The PS (Smári et al., 1994) and PSQ (Huppert et al., 2002) were validated with
clinical participants, and the studies reported positive psychometric findings relating to
internal consistency, reliability, and hypothesis testing. However, methodological problems
meant that the quality of evidence for these areas was often rated poorly. Furthermore, there
was little consideration of how appropriate these specifically ‘non-clinical’ assessments were
for a clinical population, as arguably neither questionnaire could accurately assess the range
of potential paranoid thoughts experienced by those diagnosed with schizophrenia.
Measures Assessing the Upper Paranoia Hierarchy
The PIQ (McKay et al., 2006), PaDS (Melo et al., 2009), PDI (Peters et al., 1999),
and SSPS (Freeman et al., 2007) were designed with scales to measure persecutory beliefs,
reflecting the upper levels of the paranoia hierarchy (Freeman et al., 2005). Freeman and
Garety’s (2000) definition of ‘persecutory’ was utilised in the development of items for the
PIQ, PaDS, and SSPS. Alternatively, the PDI used a definition developed by experts (Peters
et al., 1999). However, some PaDS items appear to only ‘imply’ persecutory ideas (Melo et
al., 2009); it being questionable whether items such as “There are people that think of me as a
bad person” specifically assess a perception of being at risk of harm. The PDI rates delusions
on dimensions of conviction, pre-occupation and distress, whereas the PaDS also measures
how deserved persecution is perceived to be (Trower & Chadwick, 1995). While the
‘persecution’ scale of the PaDS had some acceptable measurement properties, the properties
of the ‘deservedness’ subscale are less evidenced, due to large amounts of missing data (Melo
et al., 2009).
The PDI has items to assess PDs, alongside questions assessing other types of
delusions (e.g. grandiose; Peters et al., 1999). Statements were worded to represent
‘attenuated’ versions of delusions, appropriate for general population samples. Although the
40-item PDI was designed with four PD items, Peters et al. (1999) identified three
26
components through factor analysis which relate to ‘paranoia’, covering a broader construct
than just persecution (e.g. suspiciousness). However, Peters et al. argued that they had not
aimed to “measure a limited number of well-defined subscales… but rather to sample as wide
a variety of delusions as possible” (p. 562).
Six further studies reported PDI subscales relating to paranoia (Jung et al., 2008;
López-Ilundain et al., 2006; Peters et al., 1999; Peters et al., 2004; Prochowitz & Gaw da,
2015; Verdoux et al., 1998), with a lack of consistency in the type and number of subscales
identified. Furthermore, Jones and Fernyhough (2007) demonstrated the inadequate internal
consistency of scores on previously identified paranoia subscales of the PDI, and reported a
better fitting unidimensional factor structure, measuring general delusion-proneness.
Similarly, while Jung et al. (2008) and Prochowitz and Gaw da (2015) initially extracted
factors relating to persecution, they argued that the first underlying factor for the measure is
highly dominant and suggested that a unidimensional factor structure is preferable. Finally,
although latent class analyses using the PDI identified a ‘paranoid’ class of participants, the
‘paranoid’ items endorsed by participants were not consistent across samples (Cella et al.,
2011; Rocchi et al., 2008).
Both the SSPS (Freeman et al., 2007) and 10-item PIQ (McKay et al., 2006; Van
Dongen et al., 2011) have evidence of construct validity and test-retest reliability for their test
scores. However, they were designed as unidimensional scales, without any assessment of
structural validity (Freeman et al., 2007; McKay et al., 2006). Jones et al. (2008) did show
that scores on a 7-item PIQ had good internal consistency and better fitted a unidimensional
structure than the 10-item measure.
The SSPS assesses state persecutory ideation in the moment (as opposed to
persecutory ideas over weeks/months) and was designed for studies where paranoia is
assessed in a virtual reality (VR) environment (Freeman et al., 2007). However, there has
27
been no assessment of how responsive the scale is to momentary changes in paranoia, which
is particularly important for a state measure.
Content validity and cross-cultural validity ratings were poor for all measures
assessing the upper paranoia hierarchy, due to methodological limitations. For the PDI, factor
structures were variable and the cross-cultural validity of scores from various European
samples could not always be assessed as papers with the data needed to appraise these
analyses were not available in English (German version; Lincoln, Keller, & Rief, 2009;
Italian version; Preti, Marongiu, Petretto, Miotto, & Masala, 2002).
Most measurement properties for the persecutory measures were established with
non-clinical populations. However, the PIQ was also validated with clinical participants
(McKay et al., 2006). Construct validity hypotheses for PIQ, PaDS, and PDI were also
supported, showing significant differences in scores between clinical and non-clinical
samples (McKay et al., 2006; Melo et al., 2009), and correlations with observer-rated PDs
(Lincoln et al., 2010; McKay et al., 2006).
Measures Assessing Paranoia Spanning the Full Hierarchy
Rather than focusing upon the lower or upper paranoia hierarchy, the PC (Freeman et
al., 2005) and GPTS (Green et al., 2008) assess a range of paranoid thoughts at all levels.
Freeman et al. (2005) did not establish an a priori construct for their measure, but based upon
their findings argued that the PC assesses the hierarchy of paranoid thought, from social
evaluative concerns up to persecutory beliefs. Green et al. (2008) later used this hierarchy to
structure the GPTS item generation.
The PC assesses paranoia on dimensions of conviction and distress (Freeman et al.,
2005), and the GPTS on dimensions of preoccupation, conviction, and distress (Green et al.,
2008). Factor analyses showed that both measures have a 2-factor structure (‘persecution’ &
‘social reference’; Green et al., 2008; Ibáñez-Casas et al., 2015; ‘normal suspicions’ &
28
‘pathological delusions’; Moritz et al., 2012). For the GPTS (Green et al., 2008), factors
mapped on to the lower and higher ends of the paranoia hierarchy. However, methodologies
were rated poorly for structural validity, internal consistency, and cross-cultural validity, due
to sample size limitations (Green et al., 2008; Ibáñez-Casas et al., 2015). For the PC, some
items from the ‘pathological’ factor did not reflect extreme persecutory beliefs from the
paranoia hierarchy, and were instead described as ‘clinically relevant’ because they are
bizarre and reflect ‘first-rank’ symptoms (Moritz et al., 2012; e.g. I detected coded messages
about me in the press/TV/Radio).
The GPTS was designed for use with clinical and non-clinical participants, and
validation studies involving both groups provided some moderate evidence for its
measurement properties (e.g. reliability, hypothesis testing; Green et al., 2008; Ibáñez-Casas
et al., 2015). The PC was subsequently applied with a clinical sample, where Lincoln et al.
(2010) found a correlation between scores and observer-rated PDs. Lincoln et al. (2010) and
Moritz et al. (2012) reported that the German version of the PC has good measurement
properties. However, the cited papers were not available in English (Lincoln et al., 2009).
Furthermore, although Freeman et al. (2005) reported good internal consistency for scores on
the English PC, the unidimensionality of the scale is not evidenced, reducing the
methodological quality.
The PC has also been developed in to a state measure of paranoia (SPC; Schlier et al.,
2016); the 18 items were rephrased to ask how much they apply ‘at the moment’. Schlier et
al. (2016) generated 13-item, 5-item, and 3-item SPC scales, and demonstrated that the
shorter scales (5-item and 3-item) were more responsive to momentary changes in paranoia.
However, the data used were obtained from other studies with methodological limitations.
Furthermore, COSMIN guidance cautions that higher effect sizes do not always necessarily
indicate good responsiveness. The authors use effect sizes to demonstrate responsivity
29
without stating what the expected effect size for the interventions studied would be, making it
difficult to judge their appropriateness. Further comparison of change scores with other
measures would clarify that the SPC versions are appropriately responsive. All SPCs were
undimensional scales (although no explained variance was reported) with good internal
consistency. The 13 and 5-item measures were argued to encompass all levels of the paranoia
hierarchy, with the 3-item version having reduced content validity, but still capturing key
elements of persecutory thinking (Schier et al., 2016).
Discussion
This review aimed to critically evaluate existing self-report measures of paranoia,
based upon the constructs of paranoia that they assess and their measurement properties.
While the review identified measures developed in non-clinical populations, their
applicability to clinical samples was also considered.
Nine questionnaires were identified, assessing paranoid beliefs relating to either the
lower or upper levels of the paranoia hierarchy, or encompassing the full hierarchy (Freeman
et al., 2005). A comprehensive conceptualisation of paranoia should include thoughts relating
to varying degrees of threat and consider associated appraisals and distress. The PC (Freeman
et al., 2005) and GPTS (Green et al., 2008) were the two measures fulfilling these criteria,
capturing social reference paranoid thoughts commonly experienced across the population, as
well as persecutory beliefs common among clinical samples, and endorsed by some of the
general population. Between these measures, when combining the quality of the
methodologies of analyses and the psychometric statistics reported, the GPTS has the most
evidence for good measurement properties among clinical and non-clinical populations
(Green et al., 2008; Ibáñez-Casas et al., 2015). It also is argued to have the most clearly
defined construct underlying its items. This review therefore concludes that on the basis of
current evidence, the GPTS (Green et al., 2008) offers the most valid and informative
30
assessment of paranoia. However, some psychometric findings (e.g. internal consistency,
structural validity) require replication with a larger sample.
The PS (Fenigstein & Vanable, 1992) and PSQ (Rawlings & Freeman, 1996) were
designed to measure ‘subclinical’ paranoia (analogous to the social evaluative concerns
described by Freeman et al., 2005), as opposed to persecutory beliefs from the upper
hierarchy. However, more recent research challenges the assumption that persecutory beliefs
are always associated with psychosis, showing that they are also endorsed by some non-
clinical participants (Green et al., 2008; McKay et al., 2006). By excluding supposedly
‘extreme’ paranoid thoughts, the PS and PSQ are unable to capture the range of paranoid
experiences among a non-clinical sample, and are even less applicable for those with
psychosis, who have more persecutory beliefs.
The PDI (Peters et al., 1999), PaDS (Melo et al., 2009), PIQ (McKay et al., 2006),
and SSPS (Freeman et al., 2007) measure persecutory ideas evident in the upper paranoia
hierarchy (Freeman et al., 2005). Researchers may assess persecutory beliefs in isolation, due
to their clinical relevance. However, ideas of reference, which are not assessed by these
questionnaires, may also be clinically-relevant if they cause distress and impairment. Some of
the persecutory questionnaires do assess appraisals of beliefs, such as perceived deservedness
(PaDS; Melo et al., 2009) and conviction, pre-occupation, and distress (PDI; Peters et al.,
1999). The measurement properties of the PaDS deservedness scale (Melo et al., 2009),
however, require further validation. Furthermore, the evidence reviewed suggested that the
PDI should be used to assess general delusion proneness, rather than PDs specifically.
Although only papers reporting paranoia-related subscales were included in this review, the
use of the PDI to assess general delusion-proneness is also supported by other factor-analytic
studies (Fonseca-Pedrero, Paino, Santarén-Rosell, Lemos-Giráldez, & Muñiz, 2012; Kim et
al., 2013).
31
The PIQ (McKay et al., 2006) does not assess appraisals of persecutory ideas, but has
more evidence for acceptable measurement properties with clinical and non-clinical
populations. However, further factor analyses are required to establish whether a 10-item or
7-item measure is preferable. Given the increasing popularity of VR studies the SSPS
(Freeman et al., 2007) is also a useful tool, but requires further evaluation of its
responsiveness.
When measuring persecutory beliefs from the top of the hierarchy (Freeman et al.,
2005), prevalence rates are likely to be lower in the general population (e.g. PIQ; McKay et
al., 2006), whereas scores obtained using the PS (Feningstein & Vanable, 1992) and PSQ
(Rawlings & Freeman, 1996) may be higher. However, total scores from the latter measures
do not indicate the prevalence of paranoia specifically, as they include the assessment of
associated experiences (e.g. anger/impulsivity). Measures such as the GPTS (Green et al.,
2008) and PC (Freeman et al., 2005) therefore offer the best estimates of paranoia prevalence,
capturing the full range of potential paranoid thoughts.
The limitations of the reviewed questionnaires have implications for studies that have
used these measures. For example, by excluding the measurement of persecutory beliefs,
studies using the PS and PSQ in clinical samples (e.g. Smári et al., 1994; Craig, Hatton,
Craig, & Bentall, 2004) are unlikely to have measured a construct of paranoia appropriate for
this population. Similarly, studies identifying PDI subscales that measure specific types of
delusions, such as PDs (e.g. Jung et al., 2008), are using the measure in a way not intended by
its original authors (Peters et al., 1999). Studies using the PDI to report the prevalence of PDs
(e.g. Verdoux et al., 1998) may therefore not have assessed these experiences appropriately.
Finally, studies using the SSPS in VR settings (e.g. Freeman et al., 2015) have only assessed
persecutory thoughts, therefore not capturing potentially more commonly-occurring thoughts
from the lower paranoia hierarchy (Freeman et al., 2005). The SPC (Schlier et al., 2016) is a
32
state measure assessing a broader range of paranoid experiences, but requires further
psychometric validation (e.g. reliability, structural validity). As every measure reviewed
lacked high quality evidence for particular measurement properties, this should be considered
a limitation of all studies using self-report paranoia questionnaires.
Clinical Implications
When using paranoia questionnaires in practice, clinicians should consider that
measures assessing the full paranoia hierarchy (GPTS; Green et al., 2008; PC; Freeman et al.,
2005) will assess a greater range of service users’ experiences. Thoughts from the upper
section of the hierarchy may be experienced frequently, and thoughts from the lower
hierarchy still have potential to cause distress. Relatedly, measures assessing distress (GPTS;
Green et al., 2008; PC; Freeman et al., 2005) can highlight more troubling paranoid
experiences and evaluate distress reduction during therapy, which may be a better outcome
than reductions in thought frequency. The GPTS (Green et al., 2008) and PC (Freeman et al.,
2005) also assess appraisals of paranoid thoughts and could be used to assess the outcomes of
interventions which aim to target these (e.g. metacognitive therapy; Moritz & Woodward,
2007). Thus far, the GPTS (Green et al., 2008) has been used in randomised controlled trials
to assess the impact of various psychological interventions upon paranoia (e.g. Freeman et
al., 2017; Garety et al., 2017). However, the validity of findings from these studies could be
enhanced if there was better evidence for particular psychometric properties of the measure,
such as responsiveness to change.
The psychometric evidence for the reviewed measures suggests that when using self-
report paranoia questionnaires in practice clinicians should be mindful of limitations in their
validity. Of the measures assessing a range of paranoid thoughts, along with appraisals and
distress, the GPTS (Green et al., 2008) has the most robust psychometric evidence obtained
using clinical participants and is therefore the most recommended. The scope of this
33
questionnaire are makes it appropriate to assess paranoia among those with psychosis, and
those at risk of developing it, who may have fewer persecutory thoughts and less distress.
If the GPTS is used clinically to track change over time then collecting further data
about clinical and non-clinical norms on the scale could help to establish levels of clinically
significant change, as has been done with measures such as the CORE-OM (Barkham et al.,
2001). As there is an increasing emphasis upon the distinctiveness of different psychotic
experiences (Bentall et al., 2014), there is likely to be value in further validating paranoia-
specific questionnaire for use in interventions that specifically target these experiences.
Clinicians may wish to use measures other than the GPTS for specific purposes. If
persecutory ideas specifically are an individual’s primary difficulty, the PIQ (McKay et al.,
2006) could be used, and is the persecutory measure most validated with clinical samples.
Clinicians might also wish to assess the perceived deservedness of persecution, and could
therefore use the PaDS (Melo et al., 2009). However, they should be aware of the limitations
of using these measures, highlighted in this review.
Limitations
The questionnaires favoured in this review were based upon (GPTS; Green et al.,
2008), or resulted in (PC; Freeman et al. 2005), the development of Freeman et al.’s (2005)
paranoia hierarchy. The hierarchy is one conceptualisation of paranoia, and adopting an
alternative definition may have influenced the conclusions of the review. However, Freeman
et al.’s hierarchy is currently the most comprehensive model of paranoid cognition, with
other research often failing to distinguish paranoid and persecutory beliefs (McKay et al.,
2006). If there was a richer discussion within the academic literature about how thought
content can be defined as ‘paranoid’, this would have perhaps enhanced the appraisal of the
paranoia constructs within questionnaires. The Freeman et al. (2005) model itself could also
benefit from a more detailed definition of the levels of the hierarchy, such as a more thorough
34
discussion of what constitutes a ‘social evaluative concern’ and how this relates to paranoia.
Indeed, the lower parts of the hierarchy may be considered inappropriate in the assessment of
paranoia due to the overlap with other difficulties (e.g. anxiety). However, the prevalence of
social reference thoughts among paranoid samples (Green et al., 2008) and the strong
association between persecutory thoughts and self-consciousness (Combs & Penn, 2004;
Freeman et al., 2012; Smári et al., 1994), indicates the close relationship between the lower
and upper paranoia hierarchy. Moreover, the authors’ focus upon paranoia in the general
population indicated a need for a broad and inclusive paranoia model.
The exclusion from the review of personality and psychotic symptom measures with
paranoia subscales is a limitation, as researchers may wish to use these subscales in isolation
to assess paranoia. However, it is argued that questionnaires which are focussed purely on
paranoia and delusions are likely to have a more tightly-defined construct of paranoid
ideation, whereas broader measures may lack this. Scales specifically designed to assess
paranoia will also provide more psychometric data relevant to the assessment of paranoia,
whereas broader measures may report properties of scales that include non-paranoid items.
Within this review some measures did take items from other questionnaires assessing a range
of constructs (e.g. the PSQ included items from large inventories of psychiatric symptoms
and schizotypal personality). The inclusion of questionnaires that integrated items from
schizotypy scales may also present a direct conflict with the aim of the review; to focus upon
paranoid cognitions, as opposed to personality structures. However, this relates to a wider
discussion about whether one can distinguish an item assessing a suspicious personality trait
from an item assessing a paranoid thought.
Non-English papers were not accessed, limiting the ability to thoroughly evaluate the
cross-cultural validity of some questionnaires (e.g. the Korean PaDS; Ko & Kim, 2016; and
the Iranian GPTS; Abdolmohammadi, Mohammadzadeh, Ahmadi, & Ghadiri Sourman,

35
2016). Furthermore, unpublished papers were also not included in the review, which may
have led to some publication bias.
The search strategy used within this review resulted in papers only being included if
their study aims clearly referred to the validation of measurement properties of a
questionnaire. This strategy is recommended for systematic reviews (Terwee et al., 2011),
due to challenges identifying wider studies systematically, and to exclude studies without
specific hypotheses about reliability or validity. However, it also meant that some
psychometric data may have been missed if it was not part of the central aims of the study.
Thus, despite some of the identified measures being widely cited (e.g. the original PS paper,
Feningstein & Vanable, 1992, is cited over 500 times), often a very small proportion of these
papers were included within the review. The lack of eligible studies could be a reflection of
the small proportion of studies that provide subsequent validation of the measures, which
may indicate the need for researchers to more routinely assess the measurement properties of
paranoia questionnaires that they use.
The COSMIN protocol for systematic reviews was followed for the initial database
search procedure (Terwee et al., 2011). However, the COSMIN protocol also indicates a
second subsequent search, including the names of instruments found in the initial search,
along with terms for measurement properties and the target population. While this second
search was not completed, a citation search was instead performed and deemed satisfactory in
achieving the same outcome. All published papers that cited the original development articles
for questionnaires were thus included in this review.
Most studies included in the review were appraised poorly on particular COSMIN
items (e.g. not reporting missing data, not piloting items), meaning that properties were rated
‘fair’ or ‘poor’, even if other criteria were met at a ‘good’ or ‘excellent’ level. This masked
some of the variation between studies methodological quality. Those reviewing other self-
36
report measures have described COSMIN criteria as overly strict (Burton, Abbott, Modini, &
Touyz, 2016). However, evaluating measurement properties in accordance with gold standard
recommendations facilitated a higher quality systematic review and the methodological
weaknesses identified illustrate areas of improvement for future research. Furthermore, while
COSMIN provided a useful evaluative framework, the critique of the measures within this
paper extends beyond this by reflecting upon the implications of psychometric findings, even
if the methodologies used did not always meet the highly stringent COSMIN criteria.
Future Research
This review has highlighted a need for further validation of the existing paranoia
measures. Future studies should do this using rigorous methodologies recommended by
COSMIN (Mokkink et al., 2012), to ensure that the reviewed studies’ limitations are not
repeated. Studies could also employ IRT analyses to assess questionnaire properties. IRT
could be used to examine whether, in line with Freeman et al.’s (2005) hierarchy of paranoia,
clinical and non-clinical participants respond differently on items assessing different types of
paranoid cognition (e.g. persecutory, social reference). With regards to the development of
new paranoia questionnaires, authors should also pilot items with experts with professional
and lived experience of paranoia, to ensure that the content reflects realistic paranoid
experiences. Given arguments that it may be difficult for those experiencing paranoia to self-
report these experiences (Bell et al., 2007), when developing paranoia questionnaires in the
future it may be beneficial to consider including embedded validity indicators.
Obtaining more evidence for clinical and non-clinical norms on some of the most
psychometrically valid paranoia measures such as the GPTS (Green et al., 2008) would
increase their clinical applicability. Evaluating the measurement error of this tool would also
enable estimates of reliable clinical change to be developed. There have been increasing
efforts to design and evaluate interventions designed specifically to target paranoia or PDs
37
(e.g. Freeman et al., 2016). Increasing the quality of psychometric evidence for paranoia-
specific measures could allow the tools to be used in research evaluating such interventions,
and arguably lead to a more reliable and valid assessment of changes in paranoia.
To further investigate the value of assessing paranoid thoughts relating to varying
degrees of threat, future research could assess thoughts from lower down the paranoia
hierarchy in clinical populations. Studies could examine the distress associated with these
thoughts and compare them with persecutory beliefs, higher in the hierarchy. Furthermore,
building upon findings using observer-rated tools that non-distressing paranoid beliefs are
predictive of later paranoia-related distress and psychosis (e.g. Hanssen et al., 2005; Welham
et al., 2009), self-report questionnaires could be used longitudinally to examine the role of
frequency, content, and appraisals made about paranoid thoughts in this process. For
example, persecutory thoughts that are appraised as preoccupying and convincing may be
associated with an increased risk of later psychosis.
As indicated within the limitations section of this review, future literature reviews
could be conducted to include the paranoia subscales within broader measures of schizotypy,
personality, or psychopathology. It would be of particular interest to compare how paranoia is
conceptualised and assessed between these scales. For example, are items to assess a
paranoid cognition actually distinct from items that measure a paranoid personality trait?
Other reviews could also extend their scope to include measures that assess threats of harm to
wider society (e.g. conspiracy theories). While this review focussed upon paranoid ideation,
paranoid imagery is also prevalent among those with PDs (Schulze, Freeman, Green, &
Kuipers, 2013), and considering how best to assess this is also of interest. Furthermore,
although cognitive models conceptualise psychotic experiences as intrusions into awareness
(Morrison, 2001), no measures have assessed the process characteristics of paranoid thoughts,
and whether they do arise intrusively.

38
References
Abdolmohammadi, K., Mohammadzadeh, A., Ahmadi, E., & Ghadiri Sourman, F. (2016).
Validation of the Paranoid Thoughts Scale in the Iranian population. Qom University
of Medical Sciences Journal, 10, 47-53. doi:10.1080/13546800903378211
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental
disorders (5th ed.). Washington, DC: Author.
Barkham, M., Margison, F., Leach, C., Lucock, M., Mellor-Clark, J., Evans, C., . . . McGrath,
G. (2001). Service profiling and outcomes benchmarking using the CORE–OM:
Toward practice-based evidence in the psychological therapies. Journal of Consulting
and Clinical Psychology, 69, 184-196. doi: 10.1037/0022-006x.69.2.184
Barreto Carvalho, C., Pereira, V. M. D. P., da Motta, C., Pinto-Gouveia, J., Caldeira, S. N.,
Peixoto, E. B., & Fenigstein, A. (2014). Paranoia in the general population: A revised
version of the General Paranoia Scale for adolescents. European Scientific Journal,
10, 128-141. doi:10.1111/cp.12065
Baumstarck, K., Boyer, L., Boucekine, M., Aghababian, V., Parola, N., Lançon, C., &
Auquier, P. (2013). Self-reported quality of life measure is reliable and valid in adult
patients suffering from schizophrenia with executive impairment. Schizophrenia
Research, 147, 58–67. doi:10.1016/j.schres.2013.03.008
Beavan, V., Read, J., & Cartwright, C. (2011). The prevalence of voice-hearers in the general
population: A literature review. Journal of Mental Health, 20, 281-292.
doi:10.3109/09638237.2011.562262
Bell, M., Fiszdon, J., Richardson, R., Lysaker, P., & Bryson, G. (2007). Are self-reports valid
for schizophrenia patients with poor insight? Relationship of unawareness of illness to
psychological self-report instruments. Psychiatry Research, 151, 37–46.
doi:10.1016/j.psychres.2006.04.012
39
Ben-Porath, Y. S., & Tellegen, A. (2011). MMPI-RF (Minnesota Multiphasic Personality
Inventory-2 Restructured Form): Manual for administration, scoring, and
interpretation. Minneapolis, MN: University of Minnesota Press
Bentall, R. P., de Sousa, P., Varese, F., Wickham, S., Sitko, K., Haarmans, M., & Read, J.
(2014). From adversity to psychosis: Pathways and mechanisms from specific
adversities to specific symptoms. Social Psychiatry and Psychiatric Epidemiology,
49, 1011-1022. doi:10.1007/s00127-014-0914-0
Bentall, R. P., Rowse, G., Shryane, N., Kinderman, P., Howard, R., Blackwood, N., …
Corcoran, R. (2009). The Cognitive and Affective Structure of Paranoid Delusions.
Archives of General Psychiatry, 66, 236-247. doi:10.1001/archgenpsychiatry.2009.1
Ben-Zeev, D., Ellington, K., Swendsen, J., & Granholm, E. (2010). Examining a cognitive
model of persecutory ideation in the daily life of people with schizophrenia: A
computerized experience sampling study. Schizophrenia Bulletin, 37, 1248–1256.
doi:10.1093/schbul/sbq041
Berry, L. M., & Laskey, B. (2012). A review of obsessive intrusive thoughts in the general
population. Journal of Obsessive-Compulsive and Related Disorders, 1, 125-132.
doi:10.1016/j.jocrd.2012.02.002
Burton, A. L., Abbott, M. J., Modini, M., & Touyz, S. (2016). Psychometric evaluation of
self report measures of binge eating symptoms and related psychopathology: A
systematic review of the literature. International Journal of Eating Disorders,
49, 123-140. doi:10.1002/eat.22453
Cella, M., Sisti, D., Rocchi, M. B., & Preti, A. (2011). Delusional profiles among young
adults: A latent class analysis of delusion proneness. Psychiatry Research, 185,
97-101. doi:10.1016/j.psychres.2010.04.054
40
Cicero, D. C., & Kerns, J. G. (2010). Multidimensional factor structure of positive
schizotypy. Journal of Personality Disorders, 24, 327–343.
doi:10.1521/pedi.2010.24.3.327
Clark, D. M. & Wells, A. (1995). A cognitive model of social phobia. In R. Heimberg, M.
Liebowitz, D. A. Hope, & F. R. Schneier (Eds.), Social phobia: Diagnosis,
assessment and treatment (pp. 69–93). New York, NY: Guilford Press.
Combs, D. R., & Penn, D. L. (2004). The role of subclinical paranoia on social perception
and behavior. Schizophrenia Research, 69, 93–104. doi:10.1016/s0920-
9964(03)00051-3
Combs, D. R., Penn, D. L., & Fenigstein, A. (2002). Ethnic differences in subclinical
paranoia: An expansion of norms of the Paranoia Scale. Cultural Diversity and
Ethnic Minority Psychology, 8, 248. doi:10.1037//1099-9809.8.3.248
Craig, J. S., Hatton, C., Craig, F. B., & Bentall, R. P. (2004). Persecutory beliefs, attributions
and theory of mind: Comparison of patients with paranoid delusions, Asperger's
syndrome and healthy controls. Schizophrenia Research, 69, 29-33.
doi:10.1016/s0920-9964(03)00154-3
Dayton, C. M., & Macready, G. B. (2006). Latent class analysis in psychometrics. In C. R.
Rao & S. Sinharay (Eds.), Handbook of statistics 26 (pp. 421-446). Oxford, UK:
Elsevier.
Fenigstein, A., & Vanable, P. A. (1992). Paranoia and self-consciousness. Journal of
Personality and Social Psychology, 62, 129. doi:10.1037//0022-3514.62.1.129
Freeman, D. (2006). Delusions in the nonclinical population. Current Psychiatry
Reports, 8, 191-204. doi:10.1007/s11920-006-0023-1
Freeman, D., Dunn, G., Fowler, D., Bebbington, P., Kuipers, E., Emsley, R., … Garety, P.
(2012). Current paranoid thinking in patients with delusions: The presence of

41
cognitive-affective biases. Schizophrenia Bulletin, 39, 1281–1287.
doi:10.1093/schbul/sbs145
Freeman, D., Emsley, R., Dunn, G., Fowler, D., Bebbington, P., Kuipers, E., … Garety, P. A.
(2015). The stress of the street for patients with persecutory delusions: A test of the
symptomatic and psychological effects of going outside into a busy urban area.
Schizophrenia Bulletin, 41, 971-979. doi:10.1348/014466500163400
Freeman, D., & Garety, P. A. (2000). Comments on the content of persecutory delusions:
Does the definition need clarification? British Journal of Clinical Psychology, 39, 07-
414. doi:10.1348/014466500163400
Freeman, D., & Garety, P. A. (2014). Advances in understanding and treating persecutory
delusions: A review. Social Psychiatry and Psychiatric Epidemiology, 49, 1179–
1189. doi:10.1007/s00127-014-0928-7
Freeman, D., Garety, P. A., Bebbington, P. E., Smith, B., Rollinson, R., Fowler, D., … &
Dunn, G. (2005). Psychological investigation of the structure of paranoia in a non-
clinical population. The British Journal of Psychiatry, 186, 427-435.
doi:10.1192/bjp.186.5.427
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.
doi:10.1348/014466502760387461
Freeman, D., McManus, S., Brugha, T., Meltzer, H., Jenkins, R., & Bebbington, P. (2011).
Concomitants of paranoia in the general population. Psychological Medicine, 41, 923-
936. doi:10.1017/S0033291710001546
Freeman, D., Pugh, K., Green, C., Valmaggia, L., Dunn, G., & Garety, P. (2007). A measure
of state persecutory ideation for experimental studies. The Journal of Nervous and
Mental Disease, 195, 781-784. doi:10.1093/med:psych/9780199206315.003.0002

42
Freeman, D., Sheaves, B., Goodwin, G.M., Yu, L-M., Nickless, A., Harrison, P.J., . . . Espie,
C.A. (2017). The effects of improving sleep on mental health (OASIS): A randomised
controlled trial with mediation analysis. Lancet Psychiatry, 4, 749-758.
doi:10.1016/S2215-0366(17)30328-0
Freeman, D., Waite, F., Emsley, R., Dunn, G., Kingdon, D., Davies, L., & Fitzpatrick, R.
(2016). The efficacy of a new translational treatment for persecutory delusions: Study
protocol for a randomised controlled trial (The Feeling Safe Study). Trials, 17, 1-8.
doi:10.1186/s13063-016-1245-0
Fonseca-Pedrero, E., Paino, M., Santarén-Rosell, M., Lemos-Giráldez, S., & Muñiz, J.
(2012). Psychometric properties of the Peters et al. Delusions Inventory 21 in college
students. Comprehensive Psychiatry, 53, 893-899.
doi:10.1016/j.comppsych.2012.01.007
Garety, P. A., Ward, T., Kuipers, E., Waller, H., Hardy, A., Freeman, D., . . . Bebbington, P.
(2017). SlowMo, a digital therapy targeting reasoning in paranoia, versus treatment as
usual in the treatment of people who fear harm from others: Study protocol for a
randomised controlled trial. Trials, 18, 510-543. doi: 10.1186/s13063-017-2242-7
Green, C. E. L., Freeman, D., Kuipers, E., Bebbington, P., Fowler, D., Dunn, G., & Garety, P.
A. (2008). Measuring ideas of persecution and social reference: The Green et al.
Paranoid Thought Scales (GPTS). Psychological Medicine, 38, 101-111.
doi:10.1017/s0033291707001638
Grant, P., Green, M. J., & Mason, O. J. (2018). Models of schizotypy: The importance of
conceptual clarity. Schizophrenia Bulletin, sby012. doi:10.1093/schbul/sby012
Hanssen M., Bak, M., Bijl R., Vollebergh, W., van Os, J. (2005). The incidence and outcome
of subclinical psychotic experiences in the general population. British Journal of
Clinical Psychology, 44, 181–191. doi:10.1348/014466505x29611

43
Huppert, J. D., Smith, T. E., & Apfeldorf, W. J. (2002). Use of self-report measures of
anxiety and depression in outpatients with schizophrenia: Reliability and validity.
Journal of Psychopathology and Behavioral Assessment, 24, 275-283.
doi:10.1023/A:1020783101092
Ibáñez-Casas, I., Femia-Marzo, P., Padilla, J. L., Green, C. E., De Portugal, E., & Cervilla, J.
A. (2015). Spanish adaptation of the Green Paranoid Thought Scales. Psicothema 27,
74-81. doi:10.7334/psicothema2014.103
Johns, L. C., & van Os, J. (2001). The continuity of psychotic experiences in the general
population. Clinical Psychology Review, 21, 1125–1141. doi:10.1016/s0272-
7358(01)00103-9
Jones, S. R., & Fernyhough, C. (2007). Reliability of factorial structure of the Peters et al.
Delusions Inventory (PDI-21). Personality and Individual Differences, 43, 647-656.
doi:10.1016/j.paid.2007.01.018
Jones, S. R., Fernyhough, C., De-Wit, L., & Meins, E. (2008). A message in the medium?
Assessing the reliability of psychopathology e-questionnaires. Personality and
Individual Differences, 44, 349-359. doi:10.1016/j.paid.2007.08.014
Jung, H. Y., Chang, J. S., Yi, J. S., Hwang, S., Shin, H. K., Kim, J. H., ... & Kim, Y. S.
(2008). Measuring psychosis proneness in a nonclinical Korean population: Is
the Peters et al. Delusions Inventory useful for assessing high-risk individuals?
Comprehensive Psychiatry, 49, 202-210. doi:10.1016/j.comppsych.2007.08.011
Kean, J., & Reilly, J. (2014). Item response theory. In F. M. Hammond, J. F. Malec, T. G.
Nick & R. M. Buschbacher (Eds.), Handbook for clinical research: Design, statistics
and implementation (pp. 195-198). New York, NY: Demos Medical Publishing.
Kessler, R. C., Andrews, G., Mroczek, D., Ustun, B., & Wittchen, H. U. (1998). The World
Health Organization Composite International Diagnostic Interview short form

44
(CIDI SF). International Journal of Methods in Psychiatric Research, 7,
171-185. doi:10.1002/mpr.47
Kim, Y., Chang, J. S., Hwang, S., Yi, J. S., Cho, I. H., & Jung, H. Y. (2013). Psychometric
properties of Peters et al. Delusions Inventory-21 in adolescence. Psychiatry
Research, 207, 189-194. doi:10.1016/j.psychres.2012.09.002
Kim, S. H., Jung, H. Y., Hwang, S. S., Chang, J. S., Kim, Y., Ahn, Y. M., & Kim, Y. S.
(2010). The usefulness of a self-report questionnaire measuring auditory verbal
hallucinations. Progress in Neuro-Psychopharmacology and Biological Psychiatry,
34, 968–973. doi:10.1016/j.pnpbp.2010.05.005
Knowles, R., McCarthy-Jones, S., & Rowse, G. (2011). Grandiose delusions: A review and
theoretical integration of cognitive and affective perspectives. Clinical Psychology
Review, 31, 684–696. doi:10.1016/j.cpr.2011.02.009
Ko, H., & Kim, B. (2016). Reliability and validity of the Korean-Persecution and
Deservedness Scale. Korean Journal of Clinical Psychology, 35, 61-79.
doi:10.15842/kjcp.2016.35.1.004
Kvalseth, T. O. (1989). Note on Cohen’s kappa. Psychological Reports, 65, 223-26.
doi:10.2466/pr0.1989.65.1.223
Lenzenweger, M. F., Bennett, M. E., & Lilenfeld, L. R. (1997). The Referential Thinking
Scale as a measure of schizotypy: Scale development and initial construct
validation. Psychological Assessment, 9, 452–463. doi:10.1037/1040-3590.9.4.452
Lincoln, T. M., Keller, E., & Rief, W. (2009). Die erfassung von wahn und halluzinationen in
der normalbevölkerung: Deutsche adaptationen des Peters et al. Delusions Inventory
(PDI) und der Launay Slade Hallucination Scale (LSHS-R). Diagnostica, 55, 29-40.
doi:10.1026/0012-1924.55.1.29
Lincoln, T. M., Ziegler, M., Lüllmann, E., Müller, M. J., & Rief, W. (2010). Can delusions be
45
self-assessed? Concordance between self-and observer-rated delusions in
schizophrenia. Psychiatry Research, 178, 249-254.
doi:10.1016/j.psychres.2009.04.019
Liraud, F., Droulout, T., Parrot, M., & Verdoux, H. (2004). Agreement between self-rated
and clinically assessed symptoms in subjects with psychosis. The Journal of Nervous
and Mental Disease, 192, 352–356. doi:10.1097/01.nmd.000126702.30745.1d
López-Ilundain, J. M., Pérez-Nievas, E., Otero, M., & Mata, I. (2006). Peter's Delusions
Inventory in Spanish general population: Internal reliability, factor structure and
association with demographic variables (dimensionality of delusional ideation).
Actas Espanolas de Psiquiatria, 34, 94-104. Retrieved from
https://www.researchgate.net/publication/7225768_Peter%27s_delusions_inventory_i
n_Spanish_general_population_Internal_reliability_factor_structure_and_association_
with_demographic_variables_dimensionality_of_delusional_ideation
McKay, R., Langdon, R., & Coltheart, M. (2006). The Persecutory Ideation Questionnaire.
The Journal of Nervous and Mental Disease, 194, 628-631.
doi:10.1093/med:psych/9780199206315.003.0002
Melo, S. S., & Bentall, R. P. (2013). ‘Poor me’ versus ‘bad me’ paranoia: The association
between self beliefs and the instability of persecutory ideation. Psychology and
Psychotherapy: Theory, Research and Practice, 86, 146-163.
doi:10.1111/j.2044-8341.2011.02051.x
Melo, S., Corcoran, R., Shryane, N., & Bentall, R. P. (2009). The Persecution and
Deservedness Scale. Psychology and Psychotherapy: Theory, Research and
Practice, 82, 247-260. doi:10.1037/t20074-000
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G. (2009). Preferred reporting items for
systematic reviews and meta-analyses: The Prisma statement. Annals of Internal

46
Medicine, 151, 264-269. doi:10.1093/ptj/89.9.873
Mokkink, L., Terwee, B., Patrick, C., Alonso, D., Stratford, L., Knol, J., … Vet, W. (2010).
The COSMIN checklist for assessing the methodological quality of studies on
measurement properties of health status measurement instruments: An international
Delphi study. Quality of Life Research, 19, 539-549. doi: 10.1007/s11136-010-9606-8
Mokkink, L., Terwee, C., Gibbons, E., Stratford, P., Alonso, J., Patrick, D., . . . De Vet, H.
(2010). Inter-rater reliability of the COSMIN (COnsensus-based Standards for the
selection of health status Measurement Instruments) Checklist. Quality Of Life
Research, 19, 25. doi: 10.1186/1471-2288-10-82
Mokkink, L. B., Terwee, C. B., Patrick, D. L., Alonso, J., Stratford, P. W., Knol, D. L.,
… de Vet, H. C. W. (2012). The COSMIN checklist manual. Retrieved from
http://www.cosmin.nl/cosmin_checklist.html
Moritz, S., Van Quaquebeke, N., & Lincoln, T. M. (2012). Jumping to conclusions is
associated with paranoia but not general suspiciousness: A comparison of two
versions of the probabilistic reasoning paradigm. Schizophrenia Research and
Treatment, 2012, Article ID 384039. doi:10.1155/2012/384039
Moritz, S. & Woodward, T. S. (2007). Metacognitive training in schizophrenia: From
basic research to knowledge translation and intervention. Current Opinion in
Psychiatry, 20, 619-625. doi:10.1097/yco.0b013e3282f0b8ed
Morrison, A. P. (2001). The interpretation of intrusions in psychosis: An integrative
cognitive approach to hallucinations and delusions. Behavioural and Cognitive
Psychotherapy, 29, 257-276. doi:10.1017/s1352465801003010
Nittel, C. M., Lincoln, T. M., Lamster, F., Leube, D., Rief, W., Kircher, T., & Mehl, S.
(2018). Expressive suppression is associated with state paranoia in psychosis: An

47
experience sampling study on the association between adaptive and maladaptive
emotion regulation strategies and paranoia. British Journal of Clinical Psychology.
doi:10.1111/bjc.12174
Peters, E. R., Joseph, S. A., Day, S., & Garety, P. A. (2004). Measuring delusional
ideation: The 21-item Peters et al. Delusions Inventory (PDI). Schizophrenia
Bulletin, 30, 1005-1022. doi:10.1093/oxfordjournals.schbul.a007116
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. doi:10.1093/oxfordjournals.schbul.a033401
Preti, A., Marongiu, S., Petretto, D. R., Miotto, P., & Masala, C. (2002). Esperienze
psichiche inusuali: Studio di validazione della versione italiana del Peters et al.
Delusions Inventory. Rivista Sperimentale di Freniatria, 3-4, 81-100. Retrieved
from www.rivistafreniatria.it
Prochwicz, K., & Gaw da, Ł. (2015). The Polish version of the Peters et al. Delusions
Inventory: Factor analysis, reliability and the prevalence of delusion-like
experiences in the Polish population. Psychiatria Polska, 49, 1203-1222.
doi:10.12740/pp/42916
Rawlings, D., & Freeman, J. L. (1996). A questionnaire for the measurement of
paranoia/suspiciousness. British Journal of Clinical Psychology, 35, 451-461.
doi:10.1111/j.2044-8260.1996.tb01199.x
Rocchi, M. B., Sisti, D., Manca, S., Siddi, S., Mura, T., & Preti, A. (2008). Latent class
analysis of delusion-proneness: Exploring the latent structure of the Peters et al.
Delusions Inventory. The Journal of Nervous and Mental Disease, 196, 620-
629. doi: 10.1097/NMD.0b013e31818132a3
Selten, J.-P., Wiersma, D., & van den Bosch, R. J. (2000). Clinical predictors of discrepancy
48
between self-ratings and examiner ratings for negative symptoms. Comprehensive
Psychiatry, 41, 191–196. doi:10.1016/s0010-440x(00)90047-6
Schlier, B., Moritz, S., & Lincoln, T. M. (2016). Measuring fluctuations in paranoia:
Validity and psychometric properties of brief state versions of the Paranoia
Checklist. Psychiatry Research, 241, 323-332.
doi:10.1016/j.psychres.2016.05.002
Schulze, K., Freeman, D., Green, C., & Kuipers, E. (2013). Intrusive mental imagery in
patients with persecutory delusions. Behaviour Research and Therapy, 51, 7-14.
doi:10.1016/j.brat.2012.10.002.
Smári, J., Stefánsson, S., & Thorgilsson, H. (1994). Paranoia, self-consciousness, and
social cognition in schizophrenics. Cognitive Therapy and Research, 18, 387-
399. doi:10.1007/BF02357512
Sutton, J. L., Gould, R. L., Daley, S., Coulson, M. C., Ward, E. V., Butler, A. M., …
Howard, R. J. (2016). Psychometric properties of multicomponent tools designed to
assess frailty in older adults: A systematic review. BMC Geriatrics, 16.
doi:10.1186/s12877-016-0225-2
Stefanis, N. C., Smyrnis, N., Avramopoulos, D., Evdokimidis, I., Ntzoufras, I., & Stefanis, C.
N. (2004). Factorial composition of self-rated schizotypal traits among young males
undergoing military training. Schizophrenia Bulletin, 30, 335-350.
doi:10.1093/oxfordjournals.schbul.a00708
Terwee, C. B., de Vet, H. C. W., Prinsen, C. A. C., & Mokkink, L. B. (2011). Protocol
for systematic reviews on measurement properties. Retrieved from
http://www.cosmin.nl/Systematic%20reviews%20of%20measurement%20prope
rties.html
49
Terwee, C. B., Mokkink, L. B., Knol, D. L., Ostelo, R. W. J. G., Bouter, L. M., & de
Vet, H. C. W. (2012). Rating the methodological quality in systematic reviews
of studies on measurement properties: A scoring system for the COSMIN
checklist. Quality of Life Research, 21, 651-657. doi:10.1007/s11136-011-
9960-1
Trower, P., & Chadwick, P. (1995). Pathways to defense of the self: A theory of two
types of paranoia. Clinical Psychology: Science and Practice, 2, 263-278.
doi:10.1111/j.1468-2850.1995.tb00044.x
Van Dongen, J. D. M., Buck, N. M. L., Kool, A. M., & Van Marle, H. J. C. (2011).
Psychometric evaluation of the Dutch Persecutory Ideation Questionnaire (PIQ) and
its relation to aggression. Personality and Individual Differences, 51, 527-
531. doi:10.1016/j.paid.2011.05.014
Verdoux, H., van Os, J., Maurice-Tison, S., Gay, B., Salamon, R., & Bourgeois, M.
(1998). Is early adulthood a critical developmental stage for psychosis proneness? A
survey of delusional ideation in normal subjects. Schizophrenia Research, 29, 247-
254. doi:10.1016/s0920-9964(97)00095-9
Welham, J., Scott, J., Williams, G., Najman, J., Bor, W., O'Callaghan, M., & McGrath,
J. (2009). Emotional and behavioural antecedents of young adults who screen
positive for non-affective psychosis: A 21-year birth cohort study. Psychological
Medicine, 39, 625–634. doi:10.1017/s0033291708003760
Wigham, S., & McConachie, H. (2014). Systematic review of the properties of tools used to
measure outcomes in anxiety intervention studies for children with Autism
Spectrum Disorders. PLoS ONE, 9, e85268. doi:10.1371/journal.pone.0085268

Paranoia Research Paper

Uploaded by

Copyright:

Available Formats

Paranoia Research Paper

Uploaded by

Document Information

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

Paranoia Research Paper

Uploaded by

Copyright:

Available Formats

This is a repository copy of A systematic review of self-report measures of paranoia.

White Rose Research Online URL for this paper:

Version: Accepted Version

Paranoia can be conceptualised as consisting of a hierarchy of cognitions, ranging from

Study methodologies and measurement properties were evaluated according to COnsenus-

2012). Twenty-six studies, describing the validation of nine paranoia-related questionnaires,

hierarchy of paranoid thoughts, alongside associated dimensions such as pre-occupation,

for clinical practice and research are discussed.

Keywords: paranoia; self-report; questionnaires; validation; Psychometric; properties;

Public significance statement

Paranoia has traditionally been conceptualised as a symptom of psychosis-related

diagnoses such as ‘Schizophrenia’ (American Psychiatric Association, 2013). However,

prevalence of paranoid cognitions, ranging from approximately 2% to 42%. Data from a

large, nationally-representative UK sample has also found paranoid thinking to be associated

these associated difficulties contribute towards the development of paranoia, or are a

individuals affected by paranoid experiences, and to wider society.

cognitive biases (e.g. threat anticipation, jumping to conclusions, difficulty understanding

others’ mental states), and accompanying “emotion-related processes, such as anxiety,

within theories of schizotypy is associated with an increased vulnerability to develop

described as ‘paranoid’ and ‘persecutory’ synonymously, when the meaning of these

current/ongoing harm to oneself, enacted by an intentional perpetrator (Freeman & Garety,

to an exaggerated and “persistent misconception of one-self as the target of another’s

Persecutory thoughts and paranoid ideas of reference are distinguished somewhat

Freeman et al.’s (2005) hierarchy of paranoia provides a framework to organise these

their identified paranoid cognitions in a hierarchy according to the severity of perceived

worries, such as fears of rejection/vulnerability/the world being dangerous), followed by

Subsequent studies examining paranoid thoughts from Freeman et al.’s (2005)

comprehensive understanding of paranoid ideation (Freeman et al., 2005). Relatedly, many

Concoran, Shryane, & Bentall, 2009).

Much of our understanding of paranoia among both clinical and non-clinical

assessed, and the population to which the questionnaire is administered.

self-reporting their experiences. Furthermore, self-report assessments have additional

subsequent validation of these measures is important, to evidence the construct validity of

population, such as obsessive-intrusive thoughts (Berry & Laskey, 2012). Furthermore,

be used clinically within assessment or for outcome measurement within interventions.

In this review, we therefore aim to provide a critical appraisal of the measurement

properties of self-report measures of paranoia that were developed using non-clinical

full continua of experience through non-clinical to clinical As well as evaluating the

measurement properties of these questionnaires, we aim to use Freeman et al.’s (2005)

proposed hierarchy of perceived threat, as a framework to categorise the construct of paranoia

that is assessed by each measure. Thus, the reviewed questionnaires relate to

suspiciousness, persecutory), population for questionnaire development (e.g. general

population, non-clinical, community), instrument type (e.g. questionnaire, scale, inventory),

difficulties associated with paranoia (e.g. dementia, Parkinson’s) were excluded.

subsequent psychometric validation of the identified questionnaires. A citation search was

Inclusion and Exclusion Criteria

questionnaire rather than observer/interview-based assessments; 2. Studies must describe the

of paranoia or persecutory delusions. Furthermore, all questionnaire subscales must assess

constructs relating to paranoia or delusions. Questionnaires measuring a range of delusions

larger questionnaires either measuring non-delusional elements of psychosis (e.g. voice

journals; 10. Full articles must be available in English.

involved in paranoia, or reactions to/appraisals of paranoid experiences; 2. Questionnaires

disease, substance misuse, learning disability, paranoid personality disorder, or depression.

The COnsenus-based Standards for the selection of health-based Measurement

Instruments (COSMIN) protocol for systematic reviews of self-report questionnaires (Terwee

et al., 2011) was followed to appraise the measurement properties of questionnaires.

COSMIN definitions of measurement properties were therefore adopted, as were COSMIN

a multi-disciplinary panel of experts reached a consensus upon which properties to include

validity (includes structural validity, testing of hypotheses about related/unrelated constructs,

analyses establishing that measurement property (Terwee et al., 2012).