See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/318277532

DEVELOPMENT OF A SELF-REPORT MEASURE OF METACOGNITION: THE

METACOGNITION SELF-ASSESSMENT SCALE (MSAS). INSTRUMENT
DESCRIPTION AND FACTOR STRUCTURE

Article in Clinical Neuropsychiatry · June 2017

CITATIONS READS

0 4,269

1 author:

Antonino Carcione
Terzo Centro di Psicoterapia
136 PUBLICATIONS 5,494 CITATIONS

SEE PROFILE

All content following this page was uploaded by Antonino Carcione on 07 July 2017.

The user has requested enhancement of the downloaded file.

Clinical Neuropsychiatry (2017) 14, 3, 185-194

DEVELOPMENT OF A SELF-REPORT MEASURE OF METACOGNITION: THE METACOGNITION

SELF-ASSESSMENT SCALE (MSAS). INSTRUMENT DESCRIPTION AND FACTOR STRUCTURE

Roberto Pedone, Antonio Semerari, Ilaria Riccardi, Michele Procacci, Giuseppe Nicolò, Antonino Carcione

Abstract
Objective: Metacognition is a multi-component psychological construct, characterized by the ability to identify and
describe one’s own mental states and those of others. Evidence has been found for an association between impairments
in metacognitive abilities and poor social functioning, low quality of life, severity of psychopathology in Personality
Disorders (PDs). However, to date, there are few psychometrically validated instruments available for assessing the
different components of metacognition. A self-report questionnaire, the Metacognition Self-Assessment Scale (MSAS),
has been developed to evaluate the different functions of metacognition (Monitoring, Integration, Differentiation and
Decentration) as defined in the framework of Metacognitive Multi-Function Model (Semerari et al. 2003, 2007). The
aim of the present study is to preliminarily investigate the psychometric properties of the MSAS in a large non-clinical
sample.
Method: The MSAS was administered to 6659 people randomly recruited from the general population. Exploratory
and confirmatory factor analyses were carried out to examine the dimensionality of the MSAS.
Results: The results of the exploratory and confirmatory factor analyses revealed a good fit for a four-factor
model of metacognition and suggested that metacognition as measured by the MSAS is a multidimensional construct
consisting of one general factor with several sub-components. All Monitoring and Integration items loaded on the
first factor which appeared to correspond to self-directed reflective cognition, that we named Self-Reflectivity. The
Differentiation and Decentration items loaded on the second factor that captures the ability to distance oneself from
cognitions and evaluate them critically, we named it Critical Distance. Items related to Monitoring Others’ cognitions
constituted a separated factor, related to the ability to understand others’ minds, we named this factor Understanding
Other Minds. The results also supported the hyp othesis that metacognitive regulation (i.e. Mastery) constitutes a
separate metacognitive function, relatively independent of the metacognitive knowledge-related functions.
Conclusions: These preliminary results confirm that the MSAS has the premises to be validated as a reliable
instrument for measuring metacognition and its components. In particular, the MSAS could represent a useful and
flexible instrument for a rapid screening of metacognitive abilities in both clinical and non-clinical contexts.

Key words: metacognition, mentalization, mind-reading, assessment

Declaration of interest: the authors have nothing to disclose. This manuscript contains no actual or potential
conflict of interest on the part of any of its authors.

Roberto Pedone12, Antonio Semerari1, Ilaria Riccardi1, Michele Procacci1, Giuseppe Nicolò1, Antonino Carcione1
1
Terzo Centro di Psicoterapia Cognitiva - Scuola di Psicoterapia Cognitiva (SPC), Rome, Italy
2
Department of Psychology, Università degli Studi della Campania “Luigi Vanvitelli”, Italy

Corresponding author

Antonino Carcione
Terzo Centro di Psicoterapia Cognitiva - Via Ravenna 9/c 00161 Rome Italy
E-mail: carcione@terzocentro.it

Introduction metacognition (Bo et al. 2014, Dimaggio et al. 2007,

Dimaggio and Lysaker 2010, Gumley 2011, Semerari
In general and clinical psychology, a class of et al. 2003). In a broad definition, theory of mind
human mental activities, generally referred as “mind- (ToM) is the ability to attribute mental states, such as
reading” abilities, have been traditionally investigated beliefs, desires and intents, to oneself and others and to
by researchers from multiple theoretical backgrounds understand that others have beliefs, desires, intentions,
and research fields, each with its own lexicon and terms and perspectives that are different from one’s own
(Baron-Cohen 1995, Flavell 1976, Fonagy 1991, Frith (Premack and Woodruff 1978). Mentalization or
and Frith 2006, Frith and Happé 1999). A primary set mentalizing is the implicit or explicit perception or
of such terms and related constructs are: theory of mind interpretation of one’s or others’ actions as intentional,
(Premack and Woodruff 1978, Baron-Cohen 1995, that is, mediated by mental states or mental processes
Frith and Frith 2006), mentalization (Allen et al. 2008, (Fonagy 2002). Metacognition is a broad term that
Bateman and Fonagy 2004, Bateman et al. 2013), and can be used, among others, to define the abilities to

Submitted October 2016, Accepted February 2017

© 2017 Giovanni Fioriti Editore s.r.l. 185
 Roberto Pedone et al.

understand and reflect on one’s own and others’ mental and stable self-representation and with the creation of
states. All of these definitions are representative of positive and long lasting relationships (Dimaggio et al.
different research frameworks that are in part specific 2007a, Jørgensen 2010). In people with schizophrenia,
and in part overlap each other. At present, all of them metacognitive deterioration is strongly associated with
share, at least, two common objects that are, mental diminished social skills and with neuropsychological
states and functions that operate on them. Regarding and executive functions deficits (Lysaker et al. 2008,
the research framework we adopted to investigate 2011a, 2011b). Metacognitive impairments have been
mind-reading abilities, we refer to a functional-focused described in patients with personality disorders (PDs)
perspective model - the Metacognitive Multi-Function (Semerari et al. 2005; Dimaggio et al. 2009; Carcione
Model (MMFM) - by which an important aspect of et al. 2011), and they have been hypothesized to play
metacognition specifically refers to a set of skills a crucial role in the genesis and maintenance of PDs
intended as set of functions. Those mental functions (Bateman and Fonagy 2004, Dimaggio et al. 2007b,
allow people to identify mental states, reasoning Dimaggio and Lysaker 2010). Researches focused on
about them, and ascribing them to themselves and sample of specific categorical diagnosis of PD, describe
others, in order to regulate one’s own mental states specific metacognitive impairments in Borderline
and interpersonal relationships (Carcione et al. Personality Disorders (Semerari et al, 2005, 2015) and
1997, Semerari et al. 2003). Following its functional Avoidant Personality Disorder (Moroni et al. 2016)
operationalization, the MMFM states a set of functions, Poor metacognition has been globally linked with the
which are necessary to a) identify mental states and global severity of the PDs (Semerari et al. 2014) and
ascribing them to oneself and others on the basis with a cold, distant and non-assertive personality style
of facial expressions, somatic states, behaviour and (Spitzer et al. 2005, Inslegers et al. 2012). In contrast,
actions; b) reflect on and reason about mental states; c) high metacognitive capacities have been associated
use information about mental states to make decisions, with greater emotional regulation skills and a better
solve problems or psychological and interpersonal capacity to create stable interpersonal relationships
conflicts and cope with subjective suffering (Carcione (Bender et al. 2011). Due to the clinical relevance of the
et al. 2010). This definition highlights how the MMFM metacognition construct, there is increasing awareness
focuses primarily on the meta-cognitive operations - about the need for reliable assessment tools.
the metacognitive functions – in respect to the meta- Assessment of reflective abilities is possible through
representation that are the processed objects. From a) discourse analysis, b) interviews, c) self-report and
this point of view, the MMFM catches an aspect of d) laboratory tasks; each of these methods has pros and
metacognition that is different from Wells’s one (Wells cons. Self-report instruments are the most commonly
2000, Wells and Matthews 1994, 1996) that primarily used assessment methods (Paulhus and Vazire 2007),
considers metacognition as a set of beliefs about one’s but they have the limit that a person is called to use
own mental content that helps people to regulate their the presumed compromised skill. On the other hand,
attentive processes and that, in some cases, could induce discourse analysis and interviews are more valid, but
the maintenance of dysfunctional attentive processes they are more expensive in terms of time, training,
such as worry and rumination. In particular, the Self and resources. Several authors (Roberts et al. 2006)
regulatory executive function theory, also called the agree that a multi-method approach to assessment of
SREF model and developed by Wells and Matthews psychological constructs should be preferred as a way
(Wells and Matthews 1994, 1996), demonstrates how of controlling potential measurement artifacts and thus
metacognitive factors – i.e. , beliefs about cognition, increasing the validity of findings.
emotion, and behavior - can provoke and maintain The aim of the present study was to present the
various psychological disorders. Specifically, these development and the factorial structure investigation
metacognitions or content beliefs predispose individuals of a new MMFM-based self-report instrument, the
towards coping strategies that might be maladaptive, Metacognition Self-Assessment Scale (MSAS), which
ultimately compromising their psychological wellbeing could be used as a preliminary and fast screening tool
(Flavell 1976, Wells 2000, Wells and Purdon 1999). for the evaluation of metacognition.
The principal aspects of metacognition which the
MMFM focuses on both overlap and differ from the
definition of mentalization given by Bateman and Construct description
Fonagy (2004). Although mentalization’s definition
focuses on thinking about thinking (Fonagy 1991), it is The MSAS was developed from the MMFM
a multidimensional construct that can be considered as a (Semerari et al. 2003) and it is directly derived from two
four-intersecting-dimensions: automatic/controlled (i.e. already validated instruments based on the same model,
implicit/explicit); internally/externally based; self/other the Metacognition Assessment Scale (MAS), a rating
oriented; and cognitive/affective process (Bateman et scale for assessing metacognition in psychotherapy
al. 2013). The MMFM definition partially overlaps transcripts (Carcione et al. 2008, 2010; Semerari et al.
mentalization because it focuses on explicit cognitive 2003) and the Metacognition Assessment Interview
processes and on the self/other dimension but differs (MAI), a semi-structured clinical interview (Pellecchia
from it because, at its level of specification, it defines et al. 2015, Semerari et al. 2012).
specific functional processes that operate on mental Both MAS and MAI demonstrated acceptable levels
states. In the MMFM definition, this class of mind- of factorial validity, inter-rater agreement, internal
reading processes are constituted by different sub- validity and test–retest stability (Carcione et al. 2008;
functions that can be selectively impaired at different Dimaggio et al. 2009a, 2009b; Lysaker et al. 2005,
levels of regulation abilities (Choi-Kain and Gunderson 2010b, 2011a, 2011b; Pellecchia et al. 2015; Semerari
2008; Dimaggio et al. 2008, Semerari et al. 2007). et al. 2003, 2012) and scores on the MAS were shown
Currently, metacognition is considered crucial in several to be related to executive functions and treatment
areas of psychopathology (Dimaggio and Lysaker outcomes (Carcione et al. 2008; Lysaker et al. 2005,
2010, Gumley 2011). An impaired understanding of 2008; Semerari et al. 2005).We hypothesised that the
one’s own and others’ mental contents and processes MSAS would have a four-factor structure reflecting
seem to interfere with the development of an integrated the principal metacognitive abilities defined in the

186 Clinical Neuropsychiatry (2017) 14, 3

 The metacognition self-assessment scale (msas). Instrument description and factor structure

MMFM model as operationalized in the MAS and states. We hypothesised a similar relationship between
the MAI (Carcione et al. 2010, Semerari et al. 2003) Differentiation and Decentration. In such case, we
and consistent with single case studies (Dimaggio et consider that it is not possible to have a decentred
al. 2009a, 2009b; Semerari et al. 2003, 2005) and the perspective on others’ mental states if one is not able to
validation of MAI (Pellecchia et al. 2015, Semerari et recognise that one’s own perspective on any view about
al. 2012). each event is subjective and it is a representation rather
The MSAS assesses how people are able to than reality itself (i.e. Differentiation). 3) The third
identify their own and others’ mental states and coping factor captures understanding others’ minds dimension
with distress and interpersonal problems. The items (UOM) and 4) a fourth factor captures regulation and
are similar, in their content, to those of the MAS, control abilities (Mastery).
theoretically based on the literature on mentalization UOM and Decentration are not analogous
and attachment theories (Allen et al. 2008, Fonagy dimensions (Dimaggio et al. 2009a, Semerari et
1991, Fonagy and Target 1997), theory of mind (Baron- al. 2003); in fact, for example, it is possible to fully
Cohen et al. 1985, Premack and Woodruff 1978), understand the emotional states of a person but rigidly
metacognition (Wellman 1990; Wells 2000; Wells attribute them to the relationship he/she has with us.
and Mattews 1994, 1996) and, more generally, meta- The UOM dimension encompasses the processes that
representation (Frith 1992, Frith and Frith 2006, Frith ascribe mental states to others whereas Decentration
and Happé 1999, Sperber 2000). MSAS, such as MAS, captures the perspective – decentred or egocentric –
is divided in three main sections with grouped items from which UOM ability is exercised. We therefore
that refer to a) reflection on one’s own mental states expected that UOM and Decentration were represented
i.e. the Self domain (Understanding one’s own Mind by different factors.
- UM) b) reflection on others’ mental states i.e. the
Other domain (Understanting Others’ Mind - UOM and
Decentration, Dec) and c) coping with psychological Methods
suffering and interpersonal problems (Mastery - M)
(Carcione et al. 2010; Dimaggio et al. 2009b; Semerari Participants
et al. 2005, 2007).
Understanding One’s Own Mind (UM) is composed Potential participants responded to an advertisement;
of the following sub-functions: a) Monitoring is the they were screened and provided with information
ability to identify and define the components that about the study, then invited to provide written
make up an inner state in terms of thoughts, images, informed consent in accordance with the ‘Ethical
and emotions (Identification) and the variables related Principles of Psychologists and Code of Conduct’. The
to them (Relating Variable). b) Differentiation is the MSAS was included in a battery of questionnaires as
ability to differentiate between different classes of part of a larger-scale data survey collection project for
representation (e.g. dreams, fantasies, beliefs) and numerous studies. For the purpose of this study was
between representations and reality, recognising their submitted a brief check-list interview to control for
subjectivity. c) Integration is the ability to reflect on exclusion criteria. Individuals were excluded if they
different mental states and give a complete and coherent responded to have history of psychiatric diagnoses,
description of their components, with their evolution history of psychiatric or psychological treatments,
over time. It also relates to the ability to form a coherent history of severe brain injury and history of substance-
narrative. related disorders. None of the participants were taking
Understanding Others’ Minds (UOM) is composed psychotropic drugs, nor had they used them during
of the sub-functions: a) Monitoring, that is the ability the month preceding the study. A total of 7046 people
to recognise and define the emotions underlying others’ were randomly recruited from the general population
behaviours, expressions and actions and make plausible of Naples and its hinterland; 387 (5.5%) individuals
inferences about their thoughts. b) Decentration. It were excluded from this study, therefore 6659 met the
captures one’s ability to define others’ mental states by including criteria: 3049 men (45.8%) and 3610 women
forming hypotheses independent of (his or her) their (54.2%) aged from 18 to 75 years (M = 38.61; SD =
own perspective, mental functioning or involvement 13.97). Data were collected in two separate phases and
in the relationship, recognising their subjectivity. thus there were two sub-samples (Sample 1: n = 3459;
Mastery (M) encompasses the use of psychological 1552 (44.9%) men and 1907 (55.1%) women; age M
information to cope with problems of increasing levels = 33.83 years; SD = 13.68; Sample 2: n = 3200; 1497
of complexity; it relates to regulation and control (46.8%) men and 1703 (53.2%) women; age M = 33.09
activities. years; SD = 13.85). Both sub-samples were used in the
analyses.

Hypothesis
Measures
Following the MMFM indications and consistent
with the results of the previous study on MAI and The Metacognition Self-Assessment Scale (MSAS)
MAS, we hypothesised for the MSAS a four-factors The MSAS is an eighteen-item self-report measure
structure. More specifically, 1) The first factor specifically developed for the assessment of MMFM
captures Monitoring and Integration; 2) the second sub-functions. The MSAS is scored using a five-point
factor captures Differentiation and Decentration. We Likert scale (1 = never, 2 = rarely, 3 = sometimes, 4 =
consider that the ability to monitor one’s own inner frequently, 5 = almost always), which yields a raw score
state is a prerequisite for integration of mental states; range of 18 to 90. High scores on the MSAS indicate
in other words, one might find an individual with good better self-evaluation of metacognitive abilities than low
monitoring ability who nevertheless fails to integrate scores. The MSAS is designed to measure five abilities
mental states, but, on the contrary, it is not supposed of metacognition: 1) monitoring; 2) differentiation; 3)
to find an individual capable of integrating mental integration; 4) decentration and 5) mastery. Scores from
states who cannot monitor (his or her) their mental the five subscales are summed to give a total score that

Clinical Neuropsychiatry (2017) 14, 3 187

 Roberto Pedone et al.

represents the individual’s overall level of metacognitive and consistent with contemporary recommendations for
functioning. The five abilities are assessed as follows: test construction (Kline 1993, Clark and Watson 1995,
a) Monitoring is evaluated with six items divided APA 1999, Morey 2003), a test development team of
into two groups, depending on whether they relate to five experienced therapists, with three of them authors
monitoring of Self (see section A of the scale in the of the MMFM, examined the universe of the original
Appendix) or Others (section B). (A.1): the ability to MAS and MAI items and reworded each one in a
recognise one’s own representations such as thoughts self-report form. For each item by which the wording
and beliefs (Identification; UM_MON_ID1); (A.2): the was rising possible misunderstanding to the reader,
ability to recognise and verbalise one’s own emotions the team developed two alternative versions. This
(Identification; UM_MON_ID2); (A.3): the ability to operation led to the definition of a total of 18 items,
establish relations among the separate components of with 8 pairs of alternative items and 10 items with no
a mental state (Relating Variables; UM_MON_RV); alternative wording. Items were designed to be concise,
(B.1): the ability to recognise others’ representations unambiguous, and minimize content overlap with the
such as thoughts and beliefs (Identification; UOM_ other dimensions, and be understandable by someone
MON_ID1); (B.2): the ability to recognise and verbalise with a basic reading. The pool of alternative item pairs
others’ emotions (Identification; UOM_MON_ID2); was independently reviewed by the test development
(B.3): the ability to form ideas about what social or team, and each one was rated for content relevance
psychological factors generate to others’ mental states and quality using a 3-point ordinal scale, with 1 =
(Relating Variables; UOM_MON_RV); b) Integration unacceptable, 2 = fair, and 3 = good. These ratings were
is assessed with two items, (A.6): the ability to describe tallied, and items alternatives with higher ratings were
the cognitive and emotional aspects of an agent’s retained. At the end of the items definition process the
mental state and the temporal and social or interpersonal version of MSAS scale used in this study, consisted of
dynamics of change (UM_INT1); (A.7): the ability to eighteen item.
merge multiple potentially implausible or incoherent
mental scenarios into a fluent narrative (UM_INT2). c)
Differentiation is evaluated with two items, (A.4): the Procedures
awareness that representations are subjective and not
a perfect reflection of reality (UM_DIF1) and (A.5): After providing demographic information,
the ability to perceive that thoughts do not directly participants completed a research questionnaire booklet
influence reality, e.g. understanding that thinking about including the Metacognition Self-Assessment Scale
a catastrophe does not cause it (UM_DIF2). Having the (MSAS). All data were collected anonymously and
ability to Differentiate means that one is aware that a authorisations to collect the data were obtained before
memory could be false, and it is not an omen for the the study began.
future; that a goal will not realise simply because one
has expressed it and one’s predictions about the future Statistical analysis
are only one of many possible scenarios that may come
to pass. d) Decentration (section C) is evaluated with Data analysis was a three steps procedure. The
three items. (C.1): the ability to infer relationships aim of the first step was to explore the dimensional
among the separate components of others’ mental states structure of the MSAS. Since data were collected
and between their mental state and their behaviour into separate phases, Sample 1 and Sample 2 were
(DEC1); (C.2): the ability to recognise, define and independently analysed. Sample 1 and Sample 2
verbalise others’ cognitive inner states (DEC2); (C.3) were divided randomly in two equal portions. Scores
the ability to recognise, define and verbalise other’s from the first portion of Sample 1 and Sample 2 were
emotional inner state (DEC3). e) Mastery (section D) is independently subjected to exploratory factor analysis.
assessed in terms of the strategies which individuals use Principal Component Analysis is the most widely used
to exploit their knowledge of themselves and of others method of factor analysis among social scientists. It
to solve psychological and interpersonal problems. represents the optimal, in terms of least mean square
These strategies are divided into categories according error, scheme for reducing a set of variables to a small
to the complexity of the metacognitive operations number of linearly unrelated components. Because
involved. In ascending order of complexity these are of the algorithm on which it is based, the first axis
(D.1): dealing with a problem by voluntarily changing extracted generally accounts for most of the variance
one’s own behaviour (M1); (D.2): dealing with the and it is over-saturated. Rotation is used to mitigate this
problem through the regulation and management of effect, Oblimin rotation was used in this study because
one’s mental states, e.g. distracting oneself from ideas there were correlations among the MSAS dimensions
and emotions causing suffering (M2); (D.3): drawing (Jackson 2003). We used the KMO test and the
on one’s beliefs, evaluations or general knowledge of Bartlett’s test of sphericity to assess whether the data
one’s own mental functioning to deal with the problem were suitable for factor analysis, both suggested that
operating on underpinning (M3); (D.4) using one’s factor analysis was appropriate: KMO was greater than
own knowledge of other people’s mental functioning to 0.70 (KMO =0.92) and Bartlett’s test had a significance
manage the interpersonal dimension of a problem (M5); level less than p = 0.01.
(D.5) Mature acceptance of the limits to one’s capacity to In the second step, we used separate confirmatory
change one’s inner states and influence events (M5). The analyses of the second portion of Sample 1 and on
MSAS item scale was developed following the MMFM Sample 2 to cross-validate the structure on two samples
model implementation mentioned above. The first stage drawn from the same population. We used confirmatory
of test development was to adapt the contents of the core factor analysis to examine a model that fitted the matrix
construct dimensions from the facets developed for the of the 18 items of the MSAS and could be considered
MAS (Carcione et al. 2008, 2010; Semerari et al. 2003) conceptually and theoretically plausible.
and MAI (Pellecchia et al. 2015, Semerari et al. 2012) Multiple statistical fit indices were used to assess
in order to reformulate the entire set of facets in terms the goodness of fit of the proposed models, absolute
of self-report item. This was done for each of the five fit indices: Chi Square/d.f. ratio and Goodness of Fit
dimensions (MON, DIF, INT, DEC, M). With this aim, Index (GFI); relative fit indices: Comparative Fit Index

188 Clinical Neuropsychiatry (2017) 14, 3

 The metacognition self-assessment scale (msas). Instrument description and factor structure

(CFI) and Non-Normed Fit Index (NNFI) and a non- proportions of explained variance and cumulative
centrality based index: Root Mean Square Error of explained variance are showed in table 1.
Approximation (RMSEA). A model with a good fit In the four-factors solution each factor explained
to the data should produce consistent results on many more than 5% of the unrotated variance (see table 1) and
different indices. We used recommended statistical the whole model explained 57% of the total variance.
criteria for goodness of fit (Kline 1998, Netemeyer We consider this solution to provide interesting
et al. 2004, Ullman 1996): RMSEA < 0.06; CFI, NFI evidence about our theoretical hypotheses about the
and NNFI > 0.90 and small Expected Cross Validation organisation of MMFM functions. All UM monitoring
Index (ECVI) (Hu and Bentler 1999). Data analysis was and the integration items loaded on the first factor (F1:
carried out using Statistical Package for Social Science Self reflectivity); differentiation and decentration items
(IBM Corporation, Route 100 Somers, NY 10589, State loaded on the second factor (F2: Critical distance); all
of New York, United States, Version 15) and LISREL 8 items related to Mastery functions loaded on the third
(Joreskog et al. 2001). factor (F3: Mastery) and UOM monitoring items loaded
In the last step, the four subscales of the MSAS on the fourth factor (F4: Understanding others’ mind
were explored from Sample 1 and Sample 2, to provide monitoring).
information on the distribution of scores. Descriptive The four-factors solution demonstrated that
statistics for the MSAS dimensions and total score were the MSAS could be considered to be composed
calculated; no differences between men and women of separate and distinct components, representing
were found. The internal consistency of the scale was distinct metacognitive functions. Monitoring and
assessed by calculating Cronbach’s alphas, average integration of the Self constituted one factor (F1) and
inter-item correlations (AIICs) and corrected item-total the other metacognitive functions - differentiation-
correlations (CITCs) for the sub-scores, overall MSAS decentring (F2), mastery (F3) and UOM monitoring
score and Pearson correlations among subscales. (F4) – constituted other independent albeit associated
factors. The pattern of correlations among the four-
factors suggested that three factors (F1, F2, F4) were
Results more associated with each other (r: 0.36 – 0.40) and the
remaining factor - mastery - (F3) was weakly correlated
Exploratory factor analysis (r: 0.26 - 0.29) with the other three factors (see table 3).
The exploratory factor analysis was conducted
without imposing any restrictions on potential solutions Confirmatory factor analysis
on both samples. The screening test and the resulting
eigenvalues, together with the proportions of explained Confirmatory factor analysis of a four-factors model
variance and cumulative explained variance suggested for both samples showed that the Chi Square/d.f. Ratio
that it would be fruitful to consider a four-factors is consistent with a good fit (Sample 1, 14.83; Sample
solution. Following this, we drew on the hypothetical 2, 14.63). NNFI, NFI and CFI were for both sample
model of metacognitive structure to impose a four- above 0.90, consistent with an adequate fit as well as
factors solution. The results of the analysis were RMSEA (Sample 1, 0.064; Sample 2, 0.065). Parameter
analogous for both samples; to facilitate the reader estimates are displayed in table 2.
we reported data regarding Sample 1. Eigenvalues, The results were cross-validated in the two samples
Table 1. Results from the factor analysis of the 18-item set (four-factor solution - Sample 1)
F1 F2 F3 F4 Communalities
1 Self MON - UMIDN1 0.647 -0.198 -0.021 0.010 0.555
2 Self MON - UMIDN2 0.836 -0.031 -0.037 -0.028 0.687
3 Self MON - UMREV1 0.772 -0.089 0.001 -0.058 0.627
4 Self INT - UMINT1 0.776 0.079 0.079 0.073 0.645
5 Self INT - UMINT2 0.669 0.066 0.086 0.201 0.595
6 DIF - UMDIF1 -0.041 -0.660 -0.002 0.005 0.417
7 DIF - UMDIF2 0.035 -0.677 0.038 -0.064 0.473
8 DEC - DEC1 0.040 -0.744 -0.040 0.096 0.610
9 DEC - DEC2 0.039 -0.777 -0.026 0.058 0.645
10 DEC - DEC3 0.116 -0.630 0.016 0.017 0.482
11 M1 0.008 0.060 0.703 -0.042 0.463
12 M2 0.253 0.055 0.690 -0.073 0.580
13 M3 0.040 -0.106 0.693 0.061 0.589
14 M4 -0.087 0.007 0.657 0.148 0.470
15 M5 -0.073 -0.388 0.464 -0.004 0.430
16 Other - UOMIDN1 0.129 -0.055 -0.025 0.751 0.667
17 Other - UOMIDN2 0.006 -0.064 -0.013 0.839 0.733
18 Other - UOMREV1 -0.032 0.049 0.061 0.854 0.723
Eigenvalues 5.803 1.685 1.547 1.355
Variance (%) 32.239 9.361 8.596 7.527
Cumulative Variance (%) 32.239 41.600 50.195 57.722
Note. Extraction: Oblimin Rotation. Factor loadings higher than 0.50 are marked.

Clinical Neuropsychiatry (2017) 14, 3 189

 Roberto Pedone et al.

Table 2. Confirmatory factor analysis indexes of fit on both samples

Sample χ2 df. RMSEA ECVI CFI NFI NNFI

S1 1913.66 129 0.064 6.19 0.92 0.92 0.91

S2 1887.41 129 0.065 5.95 0.91 0.91 0.92

Note: RMSEA, Root Mean Square Error of Approximation; ECVI, Expected Cross Validation Index; CFI,
Comparative Fit Index; NFI, Normed Fit Index; NNFI, Non-Normed Fit Index

and the model had adequate values on the various et al. 2003). We tested the hypothesis that there are
indices of fit, indicating that a four-dimensional latent four different factors in the MSAS corresponding to
structure could satisfactorily represent the data. the principal components of the MMFM. Results from
MSAS descriptive statistics and scores distribution. exploratory and confirmatory factor analysis suggested
Descriptive statistics for each MSAS factor, factor that metacognition as measured by the MSAS is a
reliability and the correlation values among the factors multidimensional construct consisting of one general
for both samples are shown in table 3. factor and several sub-components and suggested that
Cronbach’s alpha ranged between 0.72 and 0.87 Mastery is a distinct function relatively independent, or
for all MSAS subscales and for overall metacognitive less associated, from the others. We are fairly confident
Table 3. Descriptive statistics, internal consistency reliability and correlations among the MSAS factors (Sample
1, n = 3459; Sample 2, n = 3200)
SAMPLE 1 Mean SD Alpha F1 F2 F3 F4

F1 - Self MON INT 4.00 0.70 0.83 —

F2 - DIF / DEC 3.92 0.73 0.78 0.40** —
F3 - Mastery 3.43 0.68 0.72 0.29** 0.28** —
F4 - Other MON 3.26 0.79 0.80 0.36** 0.26** 0.28** —

MSAS Global score 3.71 0.54 0.87

SAMPLE 2 Mean SD Alpha F1 F2 F3 F4

F1 - Self MON INT 4.01 0.70 0.83 —

F2 - DIF / DEC 3.90 0.72 0.76 0.39** —
F3 - Mastery 3.43 0.67 0.72 0.29** 0.29** —
F4 - Other MON 3.23 0.80 0.80 0.37** 0.27** 0.27** —

MSAS Global score 3.69 0.53 0.86

Note: **p< 0.01.

function as measured by total MSAS score, exceeding that our procedure (testing first EFAs on two independent
the 0.70 criterion (Clark 1995, Kline 1998, Netemeyer cross validation samples, and then CFAs on two
et al. 2004, Ullman 1996). Deleting an item did not independent cross validation samples) is both robust to
produce a substantial increase in alpha for any of the sample error and overfitting and theoretically consistent
subscales or the global scale. CITCs for all the items with our anticipations and conceptual descriptions of the
were substantially higher than 0.20 and AIICs of all targeted construct .
scales ranged between 0.35 and 0.55 (Clark 1995, Kline All Monitoring and Integration items loaded on the
1998, Netemeyer et al. 2004, Ullman 1996). Pearson’s first factor, which appeared to correspond to self-directed
correlations among the MSAS subscales were all reflective cognition: we have named this factor Self-
statistically significant. Reflectivity. As predicted Differentiation-Decentration,
Taken together, these results are consistent with the UOM, and M constituted different and relatively
hypothesis that the MSAS has a four-factors structure independent factors. More specifically, Differentiation-
reflecting the metacognitive abilities defined in the Decentration factor encompasses all the sub-functions
MMFM (Carcione et al. 2008, Semerari et al. 2003); relating to ability to distance oneself from one’s
they further suggest that three of these factors relate to thoughts: a) recognition of the representational nature of
understanding of one’s own and others’ mental states. thought; b) recognition that mental representations are
not objective images of external reality; c) recognition
that one’s own and others’ beliefs may be false and d)
Discussion recognition that others can have different points of view
and beliefs. This factor captures ability to distance oneself
The aim of this study was to investigate the factorial from cognitions and evaluate them critically, so we
structure of the MSAS, a self-report tool for the quick named it Critical Distance. Items related to Monitoring
assessment of metacognitive abilities based on the Others’ cognitions constituted the third factor, apparently
MMFM which posits that metacognition has several related to ability to understand other minds: we named
specific, relatively independent sub-functions (Semerari

190 Clinical Neuropsychiatry (2017) 14, 3

 The metacognition self-assessment scale (msas). Instrument description and factor structure

this factor Understanding Other Minds. All the results also be easily used to assess change and metacognitive
related to these first three factors are consistent with improvement during treatment and at the end of it.
data from the validation of the MAI (Pellecchia et al.
2015, Semerari et al. 2012). The results also supported
the hypothesis that metacognitive regulation (i.e. Limitations
Mastery) constitutes a separate metacognitive function, In spite of the promising findings, this study has
relatively independent of the metacognitive knowledge- a number of limitations. It is important to note that
related functions. These findings are consistent with findings based on a general population sample may not
the MMFM conceptualization of metacognition, as a be reflected in a clinical population, where factorial
general ability that emerges from the integration of semi- structure should be replicated and clinical normative
independent sub-functions (Semerari et al. 2003, 2007). parameters have yet to be established. Regarding MSAS,
In conclusion, this study has shown that 1) the MSAS further examinations of the psychometric properties of
has a good factorial validity and internal consistency in the MSAS are required; establishing the concurrent
large non-clinical cross-validated samples and 2) it is validity, test-retest reliability and predictive validity
consistent with an established model of metacognition of the scale must be a priority. There is currently no
(MMFM, Semerari et al. 2003, 2007). Administration evidence of convergent validity based on consistency
of the MSAS takes approximately 10-15 minutes, and with similar instruments and in particular, tools that
as a self-report instrument it can be used to provide a measure the same functions using other methods,
fast screening assessment of metacognitive functional Reflective Function (RF) interview and questionnaire
abilities. (Fonagy et al. 1997, 2016) or MCQ (Cartwright-
The uses of MSAS may be varied. Used formally Hatton and Wells 1997). Finally, even if the MSAS
as a self-report measure, it may act as an adjunct to the distinguishes effectively between different aspects
clinical interview (MAI) and it may provide valuable of metacognitive abilities this does not mean that it
information for the case formulation and assessment of captures general aspects of metacognitive functioning;
metacognitive abilities impairment. As showed from it is for this reason that, given also the limitations
the factorial investigations, the MSAS seems to mainly inherent in particular assessment methods, we believe
detect self-related metacognition, considering that the that research on metacognition requires the synthesis of
first factor (Self M/I) accounts for as much variance findings based on the use of several different models
as all the other factors put together. The MSAS can and measuring instruments.
APPENDIX
Table 4. Metacognition Self-Assessment Scale (MSAS)
Note. For reporting the scale, English version of the MSAS was translated by two of the authors (A. C. and R. P.).
The adequacy of the English version compared with the original Italian version was iteratively checked through
back-translation by a professional English mother-tongue translator and by the MSAS authors.

The following questionnaire regards what people think about their ability to identify and describe their thoughts,
emotions and the social relationships in which they are involved. Following the statements listed below you can
indicate your judgment on what they are descriptive of yourself. Please answer to each statement marking a cross
in the appropriate box. Thanks for your cooperation!
Almost
A RESPECT TO MYSELF, USUALLY… Never Rarely Sometimes Frequently
always
I can distinguish and differentiate my own mental
1. abilities (e.g. remembering, imagining, having
1 2 3 4 5
UM_MON_ID1 fantasies, dreaming, desiring, deciding, foreseeing
and thinking).
2. I can define, distinguish and name my own
1 2 3 4 5
UM_MON_ID 2 emotions.
3. I am aware of what are the thoughts or emotions
1 2 3 4 5
UM_MON_RV that lead my actions.
4. I am aware that what I think about myself is an idea
and not necessarily true. I realize that my opinions 1 2 3 4 5
UM_DIF1
may not be accurate and may change.

5. I am aware that what I wish or what I expect may

not be realized and that I have a limited power to 1 2 3 4 5
UM_DIF2
influence things.
6. I can clearly perceive and describe my thoughts,
1 2 3 4 5
UM_INT1 emotions and relationships in which I am involved.
7. I can describe the thread that binds my thoughts
and my emotions even when they differ from one 1 2 3 4 5
UM_INT2
moment to the next.

Clinical Neuropsychiatry (2017) 14, 3 191

 Roberto Pedone et al.

Table 4. Continue

Almost
B RESPECT TO OTHERS, USUALLY … Never Rarely Sometimes Frequently
always
I can understand and distinguish the different
1. mental activities as when they are, for example,
remembering, imagining, having fantasies, 1 2 3 4 5
UOM_MON_ID1
dreaming, desiring, deciding, foreseeing and
thinking.
2. I can identify and understand the emotions of
1 2 3 4 5
UOM_MON_ID2 people I know.
3. I can describe the thread that binds thoughts and
emotions of people I know, even when they differ 1 2 3 4 5
UOM_MON_RV
from one moment to the next.

RESPECT TO “PUT YOURSELF IN SOMEBODY Almost

C SHOES”, USUALLY …
Never Rarely Sometimes Frequently
always
I’m aware that I am not necessarily at the centre of
1. the other’s thoughts, feelings and emotions and that
other’s behaviours arise from reasons and goals that 1 2 3 4 5
DEC1
can be independent from my own perspective and
from my own involvement in the relationship.
2. I am aware that others may perceive facts and
events in a different way from me and interpret them 1 2 3 4 5
DEC2
differently.
3. I am aware that age and life experience can touch
1 2 3 4 5
DEC3 other’s thoughts, emotions and behaviour.

RESPECT TO SOLVING PROBLEMS,

Almost
D Never Rarely Sometimes Frequently
always
USUALLY …
1. I can deal with the problem voluntarily imposing or
1 2 3 4 5
M1 inhibiting a behaviour on myself.
2. I can deal with the problems voluntarily trying to
1 2 3 4 5
M2 follow my own mental order.

3. I can deal with the problems trying to challenge

or enrich my views and my beliefs on problems 1 2 3 4 5
M3
themselves.
4. When problems are related to the relationship with
the other people, I try to solve them on the basis of 1 2 3 4 5
M4
what I believe to be their mental functioning.
5. I can deal with the problems, recognizing and
accepting my limitations in managing myself and 1 2 3 4 5
M5
influencing events.

Aknowledgments Baron-Cohen S (1995). Mindblindness. An essay on Autism

and Theory of Mind. MIT Press, Cambridge, MA.
We wish to thank Dr. Giancarlo Dimaggio for initial Bateman, A, Fonagy P (2004). Psychotherapy for Borderline
suggestions for this paper. Personality Disorder: a Practical Guide. Oxford
University Press, Oxford, UK.
Bateman A, Bolton R, Fonagy P (2013). Antisocial personality
References Disorder: a Mentalizing Framework, Focus - The journal
of lifelong learning in Psychiatry XI, 2, 178-186.
American Psychological Association (1999). Standards Bo S, Abu-Akel A, Kongerslev M, Haahr UH, Bateman A
for Educational and Psychological Testing, revised. (2014). Mentalizing Mediates the Relationship Between
Aera (American Educational Research Association), Psychopathy and Type of Aggression in Schizophrenia.J
Washington, DC. Nerv Ment Dis 202, 1, 55-63.
Allen JG, Bateman AW, Fonagy P (2008). Mentalizing in Choi-Kain LW, Gunderson JG, (2008). Mentalization:
Clinical Practice. American Psychiatric Publishing Inc, ontogeny, assessment, and application in the treatment
Washington, DC.

192 Clinical Neuropsychiatry (2017) 14, 3

 The metacognition self-assessment scale (msas). Instrument description and factor structure

of borderline personality disorder. Am J Psychiatry 165, Frith C (1992). The Cognitive Neuropsychology of
1127-1135. Schizophrenia. Erlbaum, Hillsdale, NJ.
Carcione, A, Dimaggio G, Conti L, Fiore D, Nicolò G, Frith C, Frith U (2006). The neural basis of mentalizing.
Semerari A (2010). Metacognition Assessment Scale v. Neuron 50, 531-534.
4.0. Unpublished manuscript, Rome. Frith U, Happé F (1999). Theory of mind and self
Carcione A, Dimaggio G, Fiore D, Nicolò G, Procacci M, consciousness: What is it like to be autistic? Mind and
Semerari A (2008). An intensive case analysis of client Language 4, 1-22.
metacognition in a good-outcome psychotherapy: Lisa’s Kline RB (1993). The handbook of psychological testing.
case. Psychotherapy Research 18, 6, 667-676. Routledge, London.
Carcione A, Falcone M, Magnolfi G, Manaresi F (1997). Kline RB (1998). Principles and Practice of Structural
La funzione metacognitiva in psicoterapia: Scala di Equation Modeling. Guilford Press, New York.
Valutazione della Metacognizione (S.Va.M.). Psicoterapia Gumley A (2011). Metacognition, affect regulation and
9, 91-107. symptom expression: a transdiagnostic perspective.
Carcione A, Nicolò G, Pedone R, Popolo R, Conti L, Fiore Psychiatry Research 190, 72-78.
D, Procacci M, Semerari A, Dimaggio, G, (2011). Joreskog K, Sorbom D, du Toit S, du Toit M (2001).
Metacognitive mastery dysfunctions in personality LISREL 8: New Statistical Features. Scientific Software
disorder psychotherapy. Psychiatry Research 190, 1, 60- International, Lincolnwood, IL.
71. Lysaker PH, Buck KD, Carcione A, Procacci M, Salvatore G,
Cartwright-Hatton S, Wells A (1997). Beliefs about worry Nicolò G, Dimaggio G (2011c). Addressing Metacognitive
and intrusions: the Meta-Cognition Questionnaire and its Capacity in the Psychotherapy for Schizophrenia:
correlates. Journal of Anxiety Disorders 11, 279-296. A conceptual model of the key tasks and processes.
Clark LA, Watson D (1995). Constructing validity: basic Psychology and Psychotherapy: Theory, Research and
issues in objective scale development. Psychological Practice 84, 58-69.
Assessment 7, 309-319. Lysaker PH, Carcione A, Dimaggio G, Johannesen JK, Nicolò
Dimaggio G, Carcione A, Nicolò G, Conti L, Fiore D, Pedone G, Procacci M (2005). Metacognition amidst narratives
R (2009). Impaired decentration in personality disorder: of self and illness in schizophrenia: Associations with
A series of single cases analysed with the metacognition neurocognition, symptoms, insight and quality of life.
assessment scale. Clinical Psychology & Psychotherapy Acta Psychiatrica Scandinavica 112, 64-71.
16, 5, 450-462. Lysaker PH, Dimaggio G, Buck KD, Carcione A, Nicolò G
Dimaggio G, Lysaker PH (2010). Metacognition and Severe (2010). Metacognition and schizophrenia: The capacity
Adult Mental Disorders: From Basic Research to for self-reflectivity and prospective assessments of work
Treatment. Routledge, London. performance over six months. Schizophrenia Research
Dimaggio G, Lysaker PH, Carcione A, Nicolò G, Semerari 122, 124-130.
A (2008). Know yourself and you shall know the other Lysaker PH, Erickson MA, Buck B, Buck KD, Olesek K,
to a certain extent: Multiple paths of influence of self- Grant M (2011). Metacognition and social function in
reflection on mindreading Consciousness and Cognition schizophrenia: Associations over a period of five months.
17, 778-789. Cognitive Neuropsychiatry 16, 3, 241-255.
Dimaggio G, Nicolò G, Semerari A, Carcione A (2013). Lysaker PH, Kyle L, Olesek L, Warman DM, Martin JM,
Investigating the personality disorder psychotherapy Salzman AK (2011b). Metacognition in schizophrenia:
process: The roles of symptoms, quality of affects, Correlates and stability of deficits in theory of mind and
emotional dysregulation, interpersonal processes, and self-reflectivity. Psychiatry Research 190, 1, 18-22.
mentalizing. Psychotherapy Research 23, 6, 624-632. Lysaker PH, Shea AM, Buck KD, Dimaggio G, Nicolò
Dimaggio G, Semerari A, Carcione A, Nicolo G, Procacci G, Procacci M, Salvatore G, Rand KL (2010b).
M (2007). Psychotherapy of personality disorders: Metacognition as a mediator of the effects of impairments
Metacognition, states of mind, and interpersonal cycles. in neurocognition on social function in schizophrenia
Routledge, London, UK. spectrum disorders. Acta Psychiatrica Scandinavica 122,
Dimaggio G, Vanheule S, Lysaker PH, Carcione A, Nicolò G 5, 405-13.
(2009b). Impaired self-reflection in psychiatric disorders Lysaker PH, Warman DM, Dimaggio G, Procacci M, LaRocco
among adults: a proposal for the existence of a network VA, Clark LK (2008). Metacognition in schizophrenia
of semi-independent functions. Consciousness and associations with multiple assessments of executive
Cognition 18, 3, 653-664. functions. Journal of Nervous & Mental Disease 196, 5,
Flavell JH (1976). Metacognition aspects of problem solving. 384-389.
In Resnick L, The nature of intelligence. Erlbaum, Morey LC (2003). Measuring personality and
Hillsdale, NJ. psychopathology. In JA Schinka and WF Velicer
Fonagy P (1991). Thinking about thinking: Some clinical and (eds) Handbook of psychology: Research methods in
theoretical considerations in the treatment of a borderline psychology, Vol. 2, pp. 377-405. Wiley, New York.
patient. The International Journal of Psychoanalysis 72, Moroni F, Procacci M, Pellecchia G, Semerari A, Nicolò
639-656. G, Carcione A, Pedone R, Colle L (2016). Mindreading
Fonagy P, Steele M, Steele H, Target M (1997). Reflective- Dysfunction in Avoidant Personality Disorder Compared
Functioning Manual version 4.1, for applicationto Adult With Other Personality Disorders. Journal of Nervous &
Attachment Interview. University College, London. Mental Disease 204, 10, 752-757.
Fonagy P, Target M (1997). Attachment and reflective Nemiah JC, Freyberger H, Sifneos PE (1976). Alexithymia: A
function: their role in self- organization. Development View of the Psychosomatic Process. In Hill OW, Modern
and Psychopathology 9, 4, 679-700. Trends in Psychosomatic Medicine, Vol. 3, pp. 430-439.
Fonagy P, Gergely G, Jurist, Elliot L, Target M (2002), Affect Butterworths, London.
Regulation, Mentalisation and the Development of the Netemeyer RG, Bearden WO, Sharma S (2004). Scaling
Self. The Other Press. Procedures: Issues and Applications. Sage Publications,
Fonagy P, Luyten P, Moulton-Perkins A, Lee Y-W, Warren Thousand, Oaks.
F, Howard S, et al. (2016) Development and Validation Paulhus DL, Vazire S (2007). The self-report method. In
of a Self-Report Measure of Mentalizing: The Reflective Robins RW, Fraley RC, Krueger R (eds) Handbook of
Functioning Questionnaire. PLoS ONE 11, 7: e0158678. Research Methods in Personality Psychology, pp. 224-

Clinical Neuropsychiatry (2017) 14, 3 193

 Roberto Pedone et al.

239. Guilford Press, New York. Semerari A, Colle L, Pellecchia G, Carcione A, Conti L,
Pellecchia G, Moroni F, Carcione A, Colle L, Dimaggio G, Fiore D, Moroni F, Nicolò G, Procacci M, Pedone R
Nicolò G, Pedone R, Semerari A (2015). Metacognition (2015). Personality disorders and mindreading: specific
Assessment Interview: instrument description and factor impairments in patients with Borderline Personality
structure. Clinical Neuropsychiatry 12, 6, 157-165. Disorder compared to other PDs. Journal of Nervous and
Premack D, Woodruff G (1978b). Does chimpanzee have a Mental Disease 203, 8, 1-6.
theory of mind? Behavioral and Brain Sciences 1, 4, 515- Semerari A, Cucchi M, Dimaggio G, Cavadini D, Carcione A,
526. Bottelli V, Nicolò G, Pedone R, Siccardi T, D’Angerio S,
Roberts BW, Harms P, Smith JL, Wood D, Webb M (2006). Ronchi P, Maffei C, Smeraldi E (2012). The development
Using Multiple Methods in Personality Psychology. of the Metacognition Assessment Interview: Instrument
In Eid M, Diener E (eds) Handbook of Multimethod description, factor structure and reliability in a non-
Measurementin Psychology. American Psychological clinical sample. Psychiatry Research 200, 890-895.
Association, Washington, D.C., pp. 321-335. Sperber D (2000). Metarepresentations. Oxford University
Semerari A, Carcione A, Dimaggio G, Falcone M, Nicolò Press, Oxford.
G, Procacci M, Alleva G (2003). How to evaluate Ullman J, (1996). Structural Equation Modeling. In Tabachnik
metacognitive functioning in psychotherapy? The B, Fidell L (eds) Using Multivariate Statistics. Harper
metacognition assessment scale and its applications. Collins, New York, pp. 709-811.
Clinical Psychology & Psychotherapy 10, 238-261. Wellman HM, (1990). The Child’s Theory of Mind. Bradford
Semerari A, Carcione A, Dimaggio G, Nicolò G, Pedone Books. MIT Press, Cambridge, MA.
R, Procacci M (2005). Metarepresentative functions in Wells A (2000). Emotional Disorders and Metacognition.
borderline personality disorder. Journal of Personality Innovative Cognitive Therapy. John Wiley & Sons Ltd,
Disorders 19, 6, 609-710. UK.
Semerari A, Carcione A, Dimaggio G, Nicolò G, Procacci Wells A, Matthews G (1994). Attention and Emotion: a
M (2007). Understanding minds: different functions and Clinical Perpective. Lawrence Erlbaum Ass., Hillsdale,
different disorders? The contribution of psychotherapy NJ.
research. Psychotherapy Research 17, 106-119. Wells A, Matthews G (1996). Modelling cognition in
Semerari A, Colle L, Pellecchia G, Buccione I, Carcione A, emotional disorder. The S-REF model. Behaviour
Dimaggio G, Nicolò G, Procacci M, Pedone R (2014). Research and Therapy 34, 11, 881-888.
Metacognitive dysfunctions in personality disorders: Wells A, Purdon C, (1999). Metacognition and cognitive-
correlations with disorder severity and personality styles. behaviour therapy: a special issue. Clinical Psychology
Journal of Personality Disorders 28, 6, 751-766. and Psychotherapy 6, 71-72.

194 Clinical Neuropsychiatry (2017) 14, 3

View publication stats