Psychoanalytic Psychology

© 2019 American Psychological Association 2019, Vol. 1, No. 1, 000

0736-9735/19/$12.00 http://dx.doi.org/10.1037/pap0000222

Assessing Mentalization: Development and Preliminary Validation of the

Modes of Mentalization Scale
Giulia Gagliardini, PhD, and Antonello Colli, PhD
University of Urbino “Carlo Bo”

The aim of this study was to provide data on the preliminary validation of a new clinician-report measure of
mentalizing modalities, the Modes of Mentalization Scale (MMS), and to test its construct validity by using
the MMS to investigate the relationship between mentalization and clinical variables, personality pathology,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

and attachment style. A random sample of 190 therapists rated an adult patient with no psychotic symptoms
This document is copyrighted by the American Psychological Association or one of its allied publishers.

in the last 6 months using the MMS, the Clinical Questionnaire, a checklist of personality disorders (PDs), and
the Adult Attachment Questionnaire. Exploratory factor analysis provided a 5-factor solution that accounted
for 54% of the variance and represented 5 mentalizing modes: excessive certainty, concrete thinking, good
mentalization, teleological thought, and intrusive pseudomentalization. Secure attachment style was positively
predicted by good mentalization and negatively predicted by intrusive pseudomentalization; disorganized
attachment style was positively predicted by concrete thinking; dismissing attachment style was predicted by
concrete thinking; and preoccupied attachment style was predicted by teleological thought, good mentaliza-
tion, and excessive certainty about mental states. Personality disorders had clinically and empirically relevant
associations with MMS factors: good mentalization was negatively associated with schizoid PD, and intrusive
pseudomentalization was negatively associated with avoidant PD and positively associated with histrionic and
narcissistic PDs. The results did not seem to be influenced by therapists’ theoretical orientation. This study
offers preliminary evidence for the validity and reliability of the MMS, which demonstrated promising
psychometric properties. Further studies need to compare the MMS to a validated scale for the assessment of
mentalization.

Keywords: mentalization, attachment style, assessment, psychopathology, reflective function

Supplemental materials: http://dx.doi.org/10.1037/pap0000222.supp

Mentalization represents “the mental process by which an indi- a phone call). Experience is valid only when its consequences are
vidual implicitly and explicitly interprets the actions of himself apparent to all (Fonagy, Bateman, & Luyten, 2012). The physical
and others as meaningful on the basis of intentional mental states and observable dimension is dominant, and information on the
such as personal desires, needs, feelings, beliefs, and reasons” inner world is gained from the external reality.
(Bateman & Fonagy, 2004, p. xxi). Problems in reflective func- Concrete thinking, or concrete comprehension, characterizes
tioning or mentalization have been found in several psychopatho- patients who experience reality and the inner world as a whole. A
logical domains, such as personality disorders (PDs), eating dis- person who experiences reality through this modality can live
orders, depression, and so forth (Bateman, Bolton, & Fonagy, under great stress, because projections of fantasies into their ex-
2013; Petersen, Brakoulias, & Langdon, 2016; Skårderud, 2007; ternal world are felt as being real (Bateman & Fonagy, 2016):
Taubner, Kessler, Buchheim, Kächele, & Staun, 2011). Patients’ excessive reactions can be understood in light of the
As suggested by Bateman and Fonagy (2016), mentalization concreteness of their perceptions of feelings. These patients tend to
failures in adulthood can take three forms: teleological thinking,
interpret behaviors in terms of situational or physical constraints
concrete comprehension, and pseudomentalization. Teleological
rather than inner mental states, and they can adopt tautological
thought characterizes persons who recognize the presence of men-
explanations that may be based on prejudice or generalization,
tal states only when they are concretized by physical, explicit
which prevents genuine self-reflection on their mental states.
forms (e.g., a patient who recognizes a therapist’s commitment
Both teleological thought and concrete thinking are typical of
only when it is shown by the addition of psychotherapy sessions or
those situations in which the internal and external worlds are
confused. In both dimensions, internal and external reality are in a
sort of isomorphism, but in teleological thought the information
about mental states is derived from the external reality (“He
bought me flowers; therefore he loves me”), whereas in concrete
Giulia Gagliardini, PhD, and Antonello Colli, PhD, Department of
Humanities, University of Urbino “Carlo Bo.”
thought the information about mental states is derived from the
Correspondence concerning this article should be addressed to Giulia inner world of the patient (“I feel abandoned; therefore you want
Gagliardini, PhD, Department of Humanities, University of Urbino “Carlo to leave me”). The same action may have different meanings if
Bo,” via Saffi, 15, 61029 Urbino, Italy. E-mail: giulia.gagliardini@uniurb.it experienced in one of the two modalities of thought: For example,

1
 2 GAGLIARDINI AND COLLI

if a therapist unexpectedly calls a patient who is experiencing a patients: For example, a specific maker can be assessed when the
teleological modality of thought, the patient may feel the thera- interaction is overly analytic or hyperactive, meaning that the
pist’s effort to help him, whereas a patient who is in a psychic narrative is too deep, with detailed but unconvincing descriptions
equivalence mode will think that the therapist is simply driven by of the subjective reactions of self and others (e.g., “I began to see
pity and not genuinely interested. that it takes two to tango. It was a perfect collusion between the
In pseudomentalization, patients can understand and reflect on two of them. What has been called in popular psychology, you
mental states only when they are not connected with reality: know, the doormat-tyrant relationship”; Fonagy et al., 1998, p. 26).
Mentalization becomes a pure intellectual game and is not related The Reflective Function Rating Scale (RFRS; Meehan, Levy,
to real experience. The extreme consequence of this prementaliz- Reynoso, Hill, & Clarkin, 2009) represents a multi-item rating
ing representation of reality can be a dissociation of thoughts and scale for assessing RF that can be applied to a range of data
feelings from reality, up to the point where they lose their mean- sources (e.g., interviews, including but not limited to the AAI) by
ing. Psychotherapy with these patients can lead to long and com- informants such as therapists or observers rating interactions. By
plex discussions that have no connection to the genuineness of conducting a principal component factor analysis on a sample of
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

their experiences (Fonagy et al., 2012). Pseudomentalization can 49 adult patients, Meehan et al. (2009) investigated the factor
This document is copyrighted by the American Psychological Association or one of its allied publishers.

assume different forms: intrusive mentalization, in which patients structure of the scale, finding the presence of three dimensions: (a)
use their own mentalizing capacities to manipulate others; over- defensive!distorted, (b) awareness of mental states, and (c) de-
active mentalization, in which patients invest huge amounts of velopmental. This scale, however, has not been used in other
energy into thinking or talking about mental states; and destruc- studies as far as we know. The RFS and the other interview-based
tively inaccurate mentalization, in which other people’s mental measures are highly reliable; however, they are time consuming
states are denied and replaced with one’s own distorted construc- because they require therapy session transcripts or interviews for
tions. the assessment (e.g., the AAI) and long trainings to be applied
The aforementioned prementalizing modalities of thought are reliably. This restricts their application in large-scale studies and
normally experienced during childhood, are gradually abandoned limits their use in clinical contexts.
developmentally, and are substituted by a sturdy mentalization, Mentalization can also be measured through questionnaires,
which is characterized by a good tolerance of uncertainty and the such as the Reflective Functioning Questionnaire (RFQ; Fonagy et
capacity to understand and describe coherently both one’s own and al., 2016), the Mentalization Questionnaire (MZQ; Hausberg et al.,
others’ mental states. A good capacity to mentalize is also related 2012), and the Mentalization Scale (MentS; Dimitrijević, Hanak,
to an awareness that people can experience contrasting feelings Altaras Dimitrijević, & Marjanović, 2018), which can be self-
and desires; moreover, a proper mentalizing stance implies a reported by patients without being time consuming. The RFQ has
genuine curiosity about one’s own and other people’s mental shown good internal consistency and can discriminate between
states, which respects the principle of the “opaqueness” of minds clinical samples and normal controls (Fonagy et al., 2016). Factor
(i.e., the knowledge that one often cannot be sure what others are analysis showed the presence of two factors named Uncertainty
thinking or feeling without being excessively disoriented by that about Mental States (RFQ_U) and Certainty about Mental States
knowledge; Bateman & Fonagy, 2016). (RFQ_C). These two factors were significantly correlated with
borderline features, severity of depression, and impulsivity (Fon-
agy et al., 2016). Moreover lower levels of reflective function were
Assessment of Mentalization in Adults
associated with nonsuicidal self-injury behaviors (N # 253; Ba-
The growing body of theoretical works on this topic is not paired doud et al., 2015).
by an equal amount of empirical research, which may be related to The MZQ has shown good internal consistency; the scale is
some problems in assessing these prementalizing modalities of composed by four factors: Refusing Self-Reflection ($ # .68),
thought. At the present time, assessment measures of mentalization Emotional Awareness ($ # .68), Psychic Equivalence Mode ($ #
can be divided into four main categories (Bateman & Fonagy, 2016): .57), and Regulation of Affect ($ # .60). Moreover, Hausberg et
(a) interviews!narrative coding systems, (b) questionnaires, (c) ex- al. (2012) found significant differences in the MZQ scores in
perimental!observational tasks, and (d) projective measures. We do relation to attachment security, with insecurely attached patients
not consider here the projective measures, which are represented showing lower levels of mentalization than did secure subjects.
specifically by only the Projective Imagination Test (Blackshaw, The factor structure of the MentS was assessed in a sample of 288
Kinderman, Hare, & Hatton, 2001). adults and 278 college students; the scale is composed of three
One of the more used and validated narrative-based measures is factors: Other-Related Mentalization (MentS-O; $ # .77), Self-
the Reflective Functioning Scale (RFS; Fonagy, Target, Steele, & Related Mentalization (MentS-S; $ # .77), and Motivation to
Steele, 1998), which is rated on the basis of the Adult Attachment Mentalize (MentS-M; $ # .76; Dimitrijević et al., 2018). Dimi-
Interview (AAI) and shows good psychometric properties. The trijević et al. (2018) found that patients with secure attachment
RFS represents an expert rating of mentalization that is useful for scored higher on MentS than did patients with insecure attachment
empirical purposes and can provide a single, global score on a and that attachment anxiety was strongly negatively correlated
Likert scale ranging from !1 (negative reflective functioning) with self-related mentalization.
to "9 (marked reflective functioning), which in some cases can be These measures can be considered helpful and not time-
paired with an indication of the specific type of impairment in consuming assessment tools; at the same time, self-report mea-
reflective functioning (RF). Even if this assessment measure does sures in this case would be biased by the fact that patients with a
not explicitly assess prementalizing modalities of thought, raters personality disorder may not be reliable when filling out mental-
can use markers that indicate failures of mentalization in adult ization measures, because they have problems with self-awareness
 MENTALIZATION MODES AND ATTACHMENT STYLE 3

(Davidson, Obonsawin, Seils, & Patience, 2003; Huprich, Born- prementalizing factors would be positively correlated with person-
stein, & Schmitt, 2011). Moreover, patients with borderline fea- ality pathology and that patients who had more hospitalizations
tures manifest limitations of their insight into the relative disad- and/or showed more self-harming behaviors and/or suicidal at-
vantages in the capacity for cooperative relationships and a limited tempts would score higher on prementalizing modes. We also
ability to approach life in a nonimpulsive manner, which may limit hypothesized that MMS prementalizing factors would be nega-
their capacity to complete self-report measures (Morey, 2014). tively related with a good capacity to mentalize and positively
Experimental!observational tasks, such as the Reading the related to insecure attachment styles, whereas a secure attachment
Mind in the Eyes Test (Baron-Cohen, Wheelwright, Hill, Raste, & style would not be significantly related to prementalizing modal-
Plumb, 2001), are based on the recognition of mental states though ities of thought and would be related to a good capacity to
the observation of facial emotions presented to patients and have mentalize.
been used in a number of studies on the recognition of emotions by
patients with psychopathology.
Some authors, however, have criticized the assumption that the
Method
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

identification of mental states though the observation of facial

This document is copyrighted by the American Psychological Association or one of its allied publishers.

expressions can be considered on the whole as an indicator of Development of the Modes of Mentalization Scale
mentalization or theory of mind (Oakley, Brewer, Bird, & Catmur,
2016). Moreover, experimental!observational tasks share a criti- The study was approved by the University of Urbino “Carlo Bo”
cism that is common to all the aforementioned methodologies of local ethics board. The MMS is a clinician-report assessment
assessing mentalization: They are mostly focused on the explicit measure of the modes of mentalization and is written in Italian (the
sides of mentalizing and do not assess the automatic and implicit English translation of the MMS, which at the present time has not
facets of the construct. been validated yet, appears in the online supplemental materials).
In light of the aforementioned considerations, we decided to In developing the MMS, we created the first set of items (N # 50)
develop a clinician-report measure to assess prementalizing mo- by considering the following four facets of mentalizing thought as
dalities of thought. Previous studies have suggested that clinicians described by different authors (e.g., Bateman & Fonagy, 2016;
tend to make highly reliable evaluations if their observations and Fonagy et al., 2012):
inferences are quantified using psychometrically sophisticated in-
1. Teleological stance. This is characterized by items re-
struments (Blagov, Bi, Shedler, & Westen, 2012; Westen & Wein-
lated to overrelying on the external aspects and not on
berger, 2004). The assessment of mentalization from a therapist
inner mental states when interpreting behaviors. Patients
perspective has two main advantages: (a) Clinicians can also
who experience this prementalizing modality of thought
evaluate implicit and automatic mentalization by observing how
may be more interested in the practical solution of prob-
their patients interact with them and (b) considering that the
lems and on people’s actions rather than on their
therapeutic relationship tends to activate the attachment system
thoughts. Therapists tend to be more “active” with these
and stress mentalizing capacities (Bateman & Fonagy, 2016),
patients and may add sessions or provide explicit advice
clinicians can evaluate mentalization in the here and now of the
on a more frequent basis than with other patients (Bate-
interaction with the patient (“online” mentalization).
man & Fonagy, 2016).
Objectives and Hypotheses of the Current Study 2. Concrete thinking. This is characterized by items related
This study was developed with the following aims: to a sort of isomorphism of a patient’s inner and outer
world: People tend to interpret behavior on the basis of
1. Describe the development of the Modes of Mentalization physical causes or invariant characteristics. People may
Scale (MMS; Colli & Gagliardini, 2015), a new clinician- also tend to interpret reality on the basis of the current
report assessment measure of mentalization, and provide experienced emotion (especially in the case of intense
initial data on its reliability and factor structure. emotions) and to interpret the world in “good or bad”
terms (Bateman & Fonagy, 2016). This prementalizing
2. Test the construct validity of the MMS by using it to modality of thought can also be related to prejudices and
investigate the relationship between the prementalizing generalizations based on heuristics.
modalities of thought and patient and therapy variables
and attachment style as they emerge during psychother- 3. Pseudomentalization. This is characterized by items re-
apy in patients with personality disorders (PDs). lated to an overinvolvement of abstract thought and of the
cognitive facets of mentalization, with patients’ being
Concerning the factor structure of the MMS, we expected to find excessively sure of people’s thoughts and treating psy-
four factors related to the theory of mentalization that guided us in chotherapy as an intellectual game in which the affective
developing it: three corresponding to the prementalizing modali- facets of mentalization are not experienced.
ties of thought (pseudomentalization, concrete thinking, and tele-
ological stance) and one related to good mentalization. In the 4. Good mentalization. Good mentalization is characterized
analysis of the relationship between those scales and other vari- by the capacities to coherently describe mental states,
ables, we made some a priori predictions. First, we hypothesized recognize that people can feel contrasting desires and
that those prementalizing modes of thought would be clinically thoughts, and harbor a fair amount of doubt about what
coherently related to clinical variables—more specifically, that other people think or feel. A sane mentalization is man-
 4 GAGLIARDINI AND COLLI

ifested through a patient’s curiosity about the compre- pharmacotherapy. Additionally, 111 patients (62%) had a diagno-
hension of mental states and is not compromised by the sis of PD, alone or in comorbidity, following criteria of the
certainty of always knowing what is good or bad (Bate- Diagnostic and Statistical Manual of Mental Disorders (5th ed.;
man & Fonagy, 2016). DSM–5; American Psychiatric Association, 2013), 78 patients
(41%) had clinically relevant personality problems, and 185 pa-
This first set of 50 items was evaluated in terms of clarity and tients had one or more other DSM–5 diagnoses. The most common
face validity by a pool of 15 clinicians who were experienced in diagnoses were mood disorders (n # 120), anxiety disorders (n #
treating PD patients and were familiar with the concept of men- 96), and substance-related and addictive disorders (n # 57).
talization: Items were rated on a 5-point Likert scale ranging from
1 (not relevant) to 5 (very relevant) for relevance and from 1 (not
clear) to 5 (very clear) for clarity. A content validity index (CVI; Therapists
Yaghmaie, 2009) was calculated by identifying the percentage of This sample consisted of 190 Caucasian therapists (76 women;
experts who rated the item as being both relevant and clear: Items 40%) women), with a mean age of 37.3 years (SD # 10.9; range #
that had a CVI over .75 remained, and the rest were discarded
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

27– 68). Three main theoretical approaches were represented: psy-

(N # 18); the remaining items were modified, based on the
This document is copyrighted by the American Psychological Association or one of its allied publishers.

chodynamic (n # 80), cognitive!behavioral (n # 57), and

experts’ suggestions. This led to the first version of the scale, mentalization-based treatment (MBT; n # 43). Ten therapists
which was composed by 32 items and was sent to a pool of 50 reported other theoretical orientations (i.e., eclectic, systemic, and
clinicians, who used it to rate a selected patient who met our integrative). The average length of clinical experience as a psy-
inclusion criteria (at least 18 years old, had had no psychotic chotherapist was 8.8 years (SD # 10.1; range # 3–35). Seventy-
disease or psychotic symptoms in the last 6 months, and had a PD four (39%) therapists were seeing the selected patients in a private
or a clinically relevant problematic in personality). We conducted clinical practice, whereas 116 (61%) were working in public
a preliminary descriptive analysis and eliminated items with skew- mental health.
ness and kurtosis values %2, mean equal to 0 or 5, and zero
variance, as well as items that did not correlate with any other item
(N # 8). The final item list contained 24 items. These first Measures
evaluations were not included in the present study but are available For validation purposes, the following additional instruments
upon request. were used.
Adult Attachment Questionnaire. The Adult Attachment
Sampling Procedure Questionnaire (AAQ; Westen & Nakash, 2005) is a 37-item clinician-
report measure designed to assess patients’ attachment styles. It is
From the rosters of the two largest Italian associations of psy-
based on a 7-point Likert scale and codifies patients’ attachment styles
chodynamic and cognitive!behavioral psychotherapy and from
nto four different factors: secure (11 items), insecure-dismissing (nine
centers specialized in the treatment of PDs, we recruited, by
items), insecure-preoccupied (eight items), and incoherent-disorgan-
e-mail, a random sample of clinicians with at least three years of
ized (nine items). In the present study the reliability coefficients of
postpsychotherapy licensure experience. We requested that they
the AAQ ranged from .69 (moderate) for the insecure!dismissing
select a patient who was at least 18 years old, had had no psychotic
factor to .88 (good) for the secure attachment style factor. The
disorder or psychotic symptoms for at least the last six months, had
insecure!preoccupied factor and incoherent!disorganized factor
seen the therapist for a minimum of eight sessions and a maximum
showed good alphas (.78 and .72, respectively).
of 18 months, and had a PD diagnosis or a clinically relevant
Personality disorder. Following the same procedure adopted
personality problem. To minimize selection biases, we directed the
in similar studies (e.g., Betan, Heim, Zittel Conklin, & Westen,
clinicians to consult their calendars to select the last patient they
2005; Colli, Tanzilli, Dimaggio, & Lingiardi, 2014), we asked
had seen during the previous week who met the study criteria. To
clinicians to rate each randomly ordered criterion for each of the
minimize rater-dependent biases, we allowed each clinician to
DSM–5 PD diagnoses (American Psychiatric Association, 2013) as
describe only one patient. We contacted 980 clinicians, of whom
present or absent. This procedure provided both a categorical
260 (26.5%) responded that they were willing to participate. Of
diagnosis (by applying DSM–5 cutoffs) and a dimensional measure
these, 236 (24.1%) were treating a patient who met the inclusion
(number of criteria met for each disorder).
criteria and were invited to participate; 190 returned completed
Clinical questionnaire. The clinical questionnaire was con-
measures, for an overall response rate of 19.4%. The clinicians
structed ad hoc for clinicians in order to obtain general information
received no remuneration. All of the participants provided written
about them, their patients, and the therapies they used. Clinicians
informed consent.
provided basic demographic and professional data, including dis-
cipline (psychiatry or psychology), theoretical approach, hours of
Patients
work, and gender, as well as patients’ ages, and other concomitant
The sample consisted of 190 Caucasian patients (66 men; 35%), therapies (e.g., pharmacotherapy). Clinicians provided additional
with a mean age of 34.3 years (SD # 11.3; range # 18 – 65). The data on the therapies, such as length of treatments and number of
average length of treatment was 12.2 months (SD # 10.6; range # sessions. To provide a more comprehensive assessment of pa-
1–18), and 26 patients (14%) had previously had one hospitaliza- tients’ problems that may be connected to PDs and/or mentalizing
tion and 38 (20%) patients had previously had two or more. deficits, respondents were also asked to use the items of the
Forty-one patients (22%) had previously attempted suicide at Clinical Questionnaire to rate the presence or absence of a list of
least once, and 101 patients (53%) were also taking some form of clinical problems (American Psychiatric Association, 2013), such
 MENTALIZATION MODES AND ATTACHMENT STYLE 5

as dissociative symptoms, self-harming behaviors, and eating dis- nations or clichés to explain emotions and to adopt bizarre explana-
orders. Clinical Questionnaire was also used to assess trauma tions of behaviors.
history in three different dimensions: sexual violence, domestic Factor 3, good mentalization (five items; coefficient $ # .83;
violence, and physical violence. min # .00, max # 5.00; M # 2.40, SD # .99), was marked by items
indicating a good capacity to recognize and coherently describe men-
Statistical Analysis tal states, united with a curious stance toward the same and an
awareness that people can experience contrasting feelings and desires.
All analyses were conducted with SPSS Statistics 20 for Windows. The items indicate a good capacity to understand the complex nature
Before performing the statistical analyses in this sample, we tested the of mental states and their relation to behaviors and the tendency to
distribution of the data with an analysis of skewness and kurtosis spontaneously refer to mental states to interpret behaviors.
values and found that the distribution of the data for the sample is Factor 4, teleological thought (three items; coefficient $ # .77;
normal. To identify the factor structure of the MMS, we conducted an min # .00, max # 5.00; M # 3.09, SD # 1.15), indicates a tendency
exploratory factor analysis using principal axis factoring and promax to rely more on the physical manifestations of mental states (i.e.,
rotation, because we hypothesized that the factors would be interde-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

actions) rather than interpreting the world in terms of beliefs, desires,

pendent or nonorthogonal. To select the numbers of factors to extract,
This document is copyrighted by the American Psychological Association or one of its allied publishers.

or thoughts; to focus more on what people do (and not on what they

we used Kaiser’s criteria (eigenvalues &1), inspection of the scree think or feel); and to be more focused on the physical, practical
plot, percentage of variance accounted for, and parallel analysis. resolution of a problem rather than on the meanings related to the
Parallel analysis was calculated using the SPSS syntax developed by situation.
O’Connor (2000), with a generation of 1,000 random permutations Factor 5, intrusive pseudomentalization (four items; coefficient
from our data set. To create factor-based scores, we included all items $ # .67), is related to a more malign form of hyper- or pseudomen-
loading !.40 for each factor to maximize reliability (coefficient talization, indicating a tendency to intrude on and manipulate other
alpha; Stevens, 2002). We calculated the Pearson correlations be- people’s lives, in which the reflections of one’s inner world do not
tween MMS factors and the length of treatment, personality pathol- seem to be genuine. This scale’s items also indicate the tendency to
ogy, and number of previous hospitalizations; we conducted analysis use therapy as an intellectual game.
of variance (ANOVA) to assess the relationship between MMS fac-
tors and self-harming behaviors, suicidal thoughts, hospitalizations, Ruling Out the Theoretical Approach Bias
substance abuse, and suicidal behaviors. To assess the relationship
between mentalization and attachment style, we applied a blockwise An important question is the extent to which the factor structure we
multiple regression analysis. To apply the regression analysis, we found simply reflects the theoretical beliefs of participating clinicians,
tested autocorrelation and multicollinearity, yielding optimal results: particularly given that 43 clinicians in the sample reported an MBT
The Durbin-Watson Test ranged from 1.49 to 1.97, the variance orientation. To evaluate this possibility, we conducted a series of
inflation factor ranged between 1.00 and 2.53, and the tolerance exploratory factor analyses, each time excluding clinicians belonging
ranged from .40 to 1.000. We calculated the partial correlations to a specific theoretical approach (psychodynamic, cognitive–
between the number of criteria assessed for each PD for each patient behavioral, and MBT). Using the same rotation and estimation pro-
and MMS factors, each time removing the effects of all the nine other cedures, the factor analyses produced the same factor structure as did
PDs. the complete sample.
We calculated an ANOVA with Bonferroni corrections considering
the three main theoretical orientations (psychodynamic, MBT, and
Results cognitive– behavioral) present in our sample. ANOVA suggested that
MBT therapists rated significantly higher excessive certainty than did
Factor Structure of the Modes of Mentalization Scale cognitive– behavioral therapists, df: F(1, 185) # 4.10, p # .02, and
higher but not significantly higher than did psychodynamic therapists,
The exploratory factor analysis suggested the presence of five
whereas cognitive– behavioral therapists rated the good mentalization
factors that accounted for 62% of the variance (see Table 1). Measures
factor significantly higher than did the MBT group, df: F(1, 185) #
of sampling adequacy had good results (Kaiser–Meyer–Olkin # .84).
8.39, p # .000. Finally, MBT and psychodynamic therapists rated the
The intercorrelations among the five factors ranged from !.44 to .47.
teleological factor significantly higher than did cognitive– behavioral
Factor 1, excessive certainty (six items; coefficient $ # .91; min-
therapists, df: F(1, 185) # .84, p # .01.
imum [min] # .00, maximum [max] # 5.00; M # 2.80, SD # 1.12),
was marked by items indicating an overactivation of mentalization, in
Associations With Clinical and Therapy Variables
which patients show an excessive certainty about mental states and
think that they can provide all the answers regarding other people’s We calculated the Pearson correlations to investigate the relation-
inner worlds. The items of this factor indicate the tendency to be ship between MMS factors and patient and therapy variables (see
excessively sure of other people’s motivations, the inability to con- Table 2). Correlations show that there is a moderate positive correla-
sider different perspectives, and the belief that one always knows tion between MMS prementalizing factors and the number of PD
what others are thinking or feeling. criteria and a moderate negative correlation with the good mentaliza-
Factor 2, concrete thinking (six items; coefficient $ # .79; min # tion factor of the MMS and personality pathology. Moreover, the
.00, max # 5.00; M # 2.40, SD # 1.04), was marked by items number of previous hospitalizations has a small but positive correla-
indicating the tendency to interpret reality on the basis of heuristics tion with the teleological factor and a small negative correlation with
and prejudices and/or on the basis of physical or invariant constraints. the good mentalization factor. We found no significant correlation
This factor’s items describe the tendency to use commonsense expla- between MMS factors and the length of treatment.
 6 GAGLIARDINI AND COLLI

Table 1
Factor Structure of the Modes of Mentalization Scale (N # 190)

'
Factor and items 1 2 3 4 5

Factor 1. Excessive certainty

P. tends to express an excessive certainty about other people’s thoughts or feelings. .91 .03 !.03 !.15 .01
P. is excessively sure of the motivations and/or thoughts and/or emotions of others. .90 !.04 .12 .08 !.03
P. seems to have all the answers on his/her own and/or other people’s behavior. .85 !.06 !.11 .05 !.01
P. believes he/she often knows what someone else is thinking or feeling. .76 .06 .07 .01 .13
P. tends to rely in an excessive way on his/her intuitive capacity. .70 .06 015 .03 .10
P. can’t consider a point of view that differs from his/her own. .47 !.20 !.29 .27 !.06
Factor 2. Concrete thinkingb
P. tends to adopt prejudice or generalization to explain his/her own or others’
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

behavior. .19 .78 !.05 .00 !.21

P. tends to interpret behaviors in terms of physical causes (e.g. illness) and/or
This document is copyrighted by the American Psychological Association or one of its allied publishers.

stable characteristics (e.g. race, cultural background, intelligence) and/or in terms

of social external factors. .02 .68 .08 .26 !.05
P. uses commonsense explanations or clichés to explain affects or feelings. .06 .65 !.08 !.02 .03
P. seems to excessively rely on the fact that external changes can change his/her
moods. !.19 .58 .06 .11 .24
P. interprets his/her own or other people’s behavior in terms of situational or
physical constraints. !.25 .51 !.02 .17 .22
P. adopts unlikely explanations of behaviors. .07 .45 .19 !.07 .05
Factor 3. Good mentalizationc
When solicited with specific questions, P. interprets behaviors in terms of mental
states. !.05 .04 .85 .04 .03
P. is curious about the comprehension of his/her own or other people’s functioning. !.03 .05 .82 !.03 !.02
P. can describe mental states coherently. .17 !.05 .75 .03 !.12
P. understands that people can experience contrasting feelings or desires. .08 .00 .70 .01 !.07
P. spontaneously interprets behaviors in terms of mental states. .04 !.04 .51 !.22 .15
Factor 4. Teleological thoughtd
P. seems to focus more on what people do rather than on what they think or feel. .01 .25 !.01 .86 .00
P. seems to be more focused on the practical resolution of a problem rather than
on the underpinning meanings. .06 .30 !.02 .47 .03
P. seems to recognize the interest of significant others only if it is supported by
concrete actions. .03 .06 !.17 .41 .11
Factor 5. Intrusive pseudomentalizatione
P.’s reflections on his/her inner world seem to be not genuine. .07 .23 !.19 !.15 .66
P. seems to use his/her mental capacities to manipulate other people. .09 !.05 .12 .23 .65
P. seems to treat therapy as an intellectual game. .27 .15 !.07 !.16 .46
P. seems to be intrusive toward other people. .26 !.23 .08 .18 .40
Note. Item loadings greater than | .40 | are in boldface. P. # patient.
a
Eigenvalue # 7.39; variance explained # 28%. b Eigenvalue # 3.72; variance explained # 14%. c
Eigenvalue # 2.08; variance explained #
8%. d Eigenvalue # 1.63; variance explained # 6%. e Eigenvalue # 1.29; variance explained # 5%.

ANOVA was used to assess possible differences among patients violence, and we found that patients with trauma history had higher
with or without a clinical disorder diagnosis (Axis I disorders in the scores on the teleological factor, but the result was not statistically
DSM–IV–TR; American Psychiatric Association, 2000): We found no significant.
significant differences in MMS scores among patients presenting
anxiety disorders, obsessive– compulsive disorder, dissociative disor-
der, panic disorder, eating disorders, and psychosomatic disorders.
Mentalization, Personality Pathology, and
We used the same analysis to assess the relationship between MMS Attachment Style
factors and self-harming behaviors, suicidal thoughts, substance We examined the relationship between mentalization, attach-
abuse, and suicidal behaviors: Patients with self-harming behaviors ment style, and PDs by using a blockwise multiple regression
had significantly higher scores on the teleological factor, F(1, 186) # analysis to measure which of the MMS factors predicted each
14.35, p # .00, and significantly lower scores on the good mental- attachment dimension measured with the AAQ (see Table 3).
ization factor, F(1, 186) # 5.62, p # .02, than did patients without. Our results show that all attachment styles had a significant
Patients with suicidal ideations scored lower on good mentalization, relationship with at least one of the MMS factors. More specifi-
F(1, 183) # 4.10, p # .04. Higher scores on the teleological factor cally, secure attachment style was positively predicted by good
were found in patients with substance abuse, F(1, 185) # 5.24, p # mentalization and negatively predicted by intrusive pseudomen-
.02, and suicidal behaviors, F(1, 186) # 5.91, p # .02. We applied talization; dismissing attachment style was predicted by concrete
ANOVA to assess whether there were differences related to MMS thinking; and preoccupied attachment style was predicted by tele-
factors between patients with and without sexual abuse and physical ological thought, good mentalization, and excessive certainty
 MENTALIZATION MODES AND ATTACHMENT STYLE 7

Table 2 ization was negatively associated with avoidant PD and positively

Pearson Correlations of MMS Factors and Clinical Variables associated with histrionic and narcissistic PDs.
(N # 190)

Patient Discussion
No. of PD No. of Therapy The first aim of this study was to provide data on the factor
Factor criteria hospitalizations (months of treatment)
structure of the MMS. Exploratory factor analysis suggested the
Excessive certainty .25!!! .05 !.13 presence of five different factors that were conceptually coherent
Concrete thinking .37!!! .08 !.04 and in line with the theory of the multidimensional nature of the
Good mentalization !.33!!! !.16! .11 construct: excessive certainty, concrete thinking, good mentaliza-
Teleological thought .39!!! .21!! !.08
Intrusive
tion, teleological stance, and intrusive pseudomentalization. The
pseudomentalization .23!! .10 !.07 factor structure that emerged seems quite robust, with a good
internal consistency for each factor, with alpha values ranging
Note. MMS # Modes of Mentalization Scale; PD # personality disorder.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

!
p " .05. !! p " .01. !!! p " .001.
from .67 to .91, and a good differentiation between factors, with
This document is copyrighted by the American Psychological Association or one of its allied publishers.

items’ not loading strongly across multiple factors (see Table 1).
The first factor, which describes a patient’s excessive certainty
about mental states. A disorganized attachment style was posi- about knowing mental states, with an overactivation of mentalizing
tively predicted by concrete thinking. and a lack of humility in relation to the knowledge of the mental
We calculated the partial correlations for the number of criteria states of others, sounds comparable to the Certainty factor of the
satisfied for each PD, each time cutting out the effects of all other PDs RFQ (Badoud et al., 2015; Fonagy et al., 2016) and is coherent
(see Table 4). Personality disorders were significantly associated with with literature that has described this specific impairment in men-
MMS factors. Excessive certainty was positively associated with talization in patients with borderline PD (e.g., Bo, Sharp, Fonagy,
narcissistic PD and negatively associated with schizoid PD. Concrete & Kongerslev, 2017) and/or characterized by grandiosity and
thinking was positively associated with borderline, histrionic, and narcissism (Ensink, Duval, Normandin, Sharp, & Fonagy, 2018).
obsessive!compulsive PDs, whereas teleological thought predicted The second factor, concrete thinking, describes a mentalizing style
borderline, narcissistic, and obsessive!compulsive PDs. Good men- characterized by a patient’s tendency to interpret behaviors in
talization negatively predicted schizoid PD. Intrusive pseudomental- terms of heuristics and prejudices and/or on the basis of physical

Table 3
Multiple Regression Model of MMS Factors Predicting Attachment Style

Personality disorder and

factors b SE ( t p R2 corr F

Secure .57 51.27!!!

Constant .48 .28 1.73 .09
Excessive certainty .03 .06 .03 .46 .65
Concrete thinking .00 .06 .00 .01 .99
Good mentalization .82 .06 .77 13.91 .000
Teleological thought .08 .06 .09 1.31 .19
Intrusive pseudomentalization !.14 .06 !.15 !2.50 .01
Dismissing .13 6.40!!!
Constant 2.03 .36 5.69 .000
Excessive certainty .07 .07 .08 .96 .34
Concrete thinking .24 .08 .27 3.18 .002
Good mentalization !.09 .08 !.09 !1.17 .25
Teleological thought .02 .08 .02 .26 .79
Intrusive pseudomentalization .03 .07 .04 .43 .67
Preoccupied .30 17.05!!!
Constant .60 .39 1.56 .12
Excessive certainty .16 .08 .15 1.99 .05
Concrete thinking !.04 .08 !.04 !.52 .60
Good mentalization .25 .08 .21 2.98 .003
Teleological thought .52 .08 .52 6.24 .000
Intrusive pseudomentalization .08 .08 .09 1.09 .28
Disorganized .14 7.37!!!
Constant 1.65 .36 4.56 .000
Excessive certainty !.04 .07 !.05 !.55 .59
Concrete thinking .27 .08 .29 3.50 .001
Good mentalization !.10 .08 !.11 !1.41 .16
Teleological thought .09 .08 .10 1.09 .28
Intrusive pseudomentalization .02 .07 .02 .20 .84
Note. MMS # Modes of Mentalization Scale; corr # corrected.
!!!
p " .001.
 8 GAGLIARDINI AND COLLI

Table 4
Partial Correlations of MMS Factors and Personality Disorders

Excessive Concrete Good Teleological Intrusive

Personality disorder certainty thinking mentalization thought pseudomentalization

Paranoid .11 !.08 !.07 .02 !.11

Schizoid !.16! .10 !.27!!! .06 .01
Schizotypal .10 .08 !.09 !.02 .09
Antisocial .09 .00 !.10 !.09 !.02
Borderline !.07 .20!! !.14 .28!!! .00
Histrionic .01 .15! .01 .01 .20!!
Narcissistic .31!!! .10 !.03 .22!! .40!!!
Dependent !.14 !.01 !.01 !.08 !.05
Avoidant .04 !.06 .13 .04 !.18!
Obsessive!compulsive .12 .25!!! !.03 .16! .14
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Note. Analysis is based on the number of criteria satisfied for each patient for each personality disorder.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

MMS # Modes of Mentalization Scale.

!
p " .05. !! p " .01. !!! p " .001.

constraints, using common sense, clichés, or even bizarre expla- may be because cognitive– behavioral therapies are more focused
nations to understand behaviors, and this is coherent with clinical on behaviors and on the external manifestations of mental states. In
literature that has described the specific difficulty some patients the future it would be important to address these issues to under-
have in using mental states to interpret behaviors (Bateman & stand something more about these differences.
Fonagy, 2016). The second aim of our study was to assess criterion validity in
The third factor, good mentalization, describes different facets relation to certain clinical variables. We found several clinically
of good reflective function such as curiosity and humility about coherent correlations between personality pathology and MMS
knowing one’s own and others’ mental states and a tendency to factors, indicating, for example, that patients with a higher number
answer in terms of mental states when solicited for answers to of PD criteria have higher scores on all the prementalizing factors.
specific demand questions but also a tendency to spontaneously This result is in line with the clinical and empirical literature on the
think in terms of mental states. The last two factors that emerged topic, which has enlightened the relationship between mentalizing
were the teleological and the intrusive mentalization. The former is problematics and personality pathology (Bateman & Fonagy,
coherent, with clinical descriptions of patients (especially trauma- 2016).
tized and borderline personality disorder [BPD] patients) who tend The teleological factor was also associated with a higher number
to rely on the physical consequences of mental states rather than on of previous hospitalizations; moreover, ANOVA showed that
their own and others’ inner worlds (Bateman & Fonagy, 2016); the higher scores on the teleological factors were present in patients
latter is in line with descriptions of mentalization in patients with with self-harming behaviors, substance abuse, and suicidal behav-
antisocial personality disorder (Bateman et al., 2013). iors. This result is in line with studies that have shown that lower
We conducted a series of exploratory factor analysis, each time levels of mentalization are associated with nonsuicidal self-injury
excluding clinicians belonging to a specific theoretical approach (Badoud et al., 2015) and impulsivity (Fonagy et al., 2016). Pa-
(psychodynamic, cognitive– behavioral, and MBT), and results tients with higher scores on teleological thought may be more
suggested that the factor structure does not seem to be affected by focused on the physical outcomes of mental states, and this may be
clinicians’ theoretical orientation. At the same time, we found related to the tendency to look for the manifestations of mental
some significant differences in relation to the excessive certainty, states in the outer reality. When this is not paired with a robust
good mentalization, and teleological factors among the three the- capacity to mentalize, patients may be more prone to acting out
oretical orientations. These results may be related to the differ- and to at-risk behaviors, because the feelings that cannot be ex-
ences among the patients in the three subsamples, but may also be pressed in words have to be acted out in the external world in order
related to theoretical differences between the different therapists. to be felt as “real.”
Specifically, the fact that MBT therapists tend to rate scores on the Contrary to our expectations, even though patients with trauma
excessive certainty factor higher, whereas cognitive– behavioral history had more problematics in mentalization, this result was not
therapists tend to provide lower scores on the good mentalization statistically significant. This may be related to the influence of
factor, may be related to the fact that MBT therapists are trained to different variables that may mediate the relationship between
recognize the less genuine forms of mentalization. Moreover, trauma history and mentalization. Moreover, one must consider
cognitive– behavioral therapists may be more focused on the cog- that the evaluations constituting this sample belong to patients who
nitive (vs. affective) processes, and this may be lead to rating a were, at the time of the assessment, in an ongoing psychotherapy;
statement from a patient as “good mentalizing communication,” these issues may have emerged throughout the treatment and have
whereas the same communication may be considered an expres- been addressed by the therapeutic dyad, with good consequences
sion of the defense mechanism of “rationalization” by a psychody- for the possibility to mentalize these events.
namic psychotherapist and as “pseudomentalizing” by an MBT We used the MMS to examine the relationship between men-
therapist. The data also suggest that MBT and psychodynamic talization and attachment style and found that impairments in
therapists assess scores on the teleological dimension higher; this mentalization were positively related to insecure attachment and
 MENTALIZATION MODES AND ATTACHMENT STYLE 9

negatively related to secure attachment. Secure attachment style methods based on the evaluations provided by external raters may
was positively predicted by good mentalization and negatively not be able to catch the more implicit and automatic facets of
predicted by intrusive pseudomentalization and excessive certainty mentalization as they manifest in the immediacy of the interaction
about mental states, whereas disorganized attachment style was with the therapist.
positively predicted by excessive certainty and negatively pre- An important limitation of this study is related to the lack of
dicted by good mentalization (see Table 3). Dismissing attachment comparison between the MMS and one of the more important and
style was predicted by teleological thought, and preoccupied at- widely used assessment measures of mentalization, the RFS. Com-
tachment style was predicted by concrete thinking and excessive paring our results with the assessment of patients’ reflective func-
certainty about mental states. These results are in line with those in tion would be essential to confirm the findings of this study. In this
the empirical literature on mentalization (Dimitrijević et al., 2018), study, the factor structure and criterion validity of the scale have
which has enlightened how attachment anxiety negatively corre- been investigated in a sample of 191 patients. Because the measure
lates with the capacity to mentalize toward the self. These subjects, has a relatively low number of items and a relatively low number
who have more problems with mentalizing toward themselves, of components, the sample size may be considered adequate, but it
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

may be more compelled by the hyperactivation of mentalization is still a small sample, and this work should be replicated including
This document is copyrighted by the American Psychological Association or one of its allied publishers.

toward others, and this may invalidate their capacity to correctly more subjects.
interpret their mental states. Despite the limitations described and the need for further vali-
Good mentalization positively predicted a secure attachment dation of the scale, our preliminary results suggest that the MMS
style and negatively predicted a disorganized attachment style but can be a reliable measure for the evaluation of mentalization, with
did not predict other insecure attachment styles. The fact that only the advantage of being economical and able to provide an articu-
disorganized attachment style is negatively related to good men- late and complex description of patients’ mentalizing capacities,
talization may indicate that disorganized patients are characterized which could be useful in everyday clinical practice and for re-
by a more severe impairment of mentalization and cannot rely on search purposes.
a good capacity to mentalize, but insecure patients may show a less
drastic impairment of this capacity. 摘要
Whereas the effect size predicting secure attachment is strong,
for the insecure types the effect size is more modest (see Table 3).
This result may indicate that, although good mentalization plays a 摘要:本研究的目标是为一个新的心理化模式的临床测量表,即”心
crucial role in the prediction of secure attachment style and the 理化量表模式”(MMS)的初步验证提供数据,并检验其构想的有效性,通
relationship between these two variables may be clearer and more 过MMS研究心理化和临床变量, 人格病理学和依恋模式之间的关系.
方法:随机样本是190位治疗师与各自的一位最近六个月里没有精神病
linear, the path that leads to insecure attachment styles may be
症状的成年患者,使用了MMS, 临床调查问卷, 人格障碍检查表和成人
more complex, and more needs to be known about other variables
依恋调查表. 结果:探索性因素分析提供了一个五因素的解答,54%的方
that may influence this relationship. 差,代表了五种心理化模式:过度确定性, 具体思维, 良好的心理化,
We also used the MMS to examine the relationship between 目的论的思考, 侵入性的假性心理化. 安全型依恋由良好的心理化给出
mentalization and personality disorders (see Table 4) and found 了一个正向预测,而侵入性的假性心理化则给出了一个负向预测;混
different specific patterns of mentalization failures. Narcissistic 乱型依恋由具体思维给出了一个正向预测;疏离型依恋由具体思维给
PD, for example, was characterized by a combination of intrusive 出预测;焦虑矛盾型依恋则由目的论的思考, 良好的心理化和精神状态
pseudomentalization, teleological thought, and excessive certainty 的过度确定性给出预测. 人格障碍与MMS因素有临床的和经验性的相
about mental states. Borderline PD was associated with concrete 关:良好的心理化与类分裂性人格障碍呈负相关,侵入性假性心理化
thinking and teleological thought, and this result is in line with the 与回避性人格障碍呈负相关,与戏剧性人格障碍和自恋性人格障碍呈
正相关. 结果似乎并没有受到治疗师的理论取向所影响. 结论:此研
clinical literature on borderline patients and these patients’ ten-
究为MMS 的信度和效度提供了初步证据,证实了其有前途的心理测量
dency to act out (Bateman & Fonagy, 2016). This result is also in
性能. 进一步的研究需要将MMS 与一个经过验证的心理化评估量表进
line with the theoretical literature on mentalization and enactments 行比较.
with BPD patients (Bateman & Fonagy, 2016) The strong negative
association of good mentalization over schizoid PD may suggest 关键词: 心理化, 依恋类型, 评估, 精神病理学, 反思功能
that the difficulties in mentalizing may be more severe in these
subjects. Our results are in line with the clinical, empirical, and
theoretical literature on PDs and mentalization. References
This study has some limitations. First, clinicians provided data American Psychiatric Association. (2000). Diagnostic and statistical man-
on patients’ mentalization and attachment styles, and this might be ual of mental disorders (4th ed., text rev.). Washington, DC: Author.
related to a bias in our results. We wanted to test the validity of the American Psychiatric Association. (2013). Diagnostic and statistical man-
scale as a first, and this began with clinicians alone, because they ual of mental disorders (5th ed.). Arlington, VA: American Psychiatric
are the selected raters of the scale. The decision to use therapists Publishing.
as assessors was also motivated by the awareness that clinicians Badoud, D., Luyten, P., Fonseca-Pedrero, E., Eliez, S., Fonagy, P., &
Debbané, M. (2015). The French version of the Reflective Functioning
can provide highly reliable and valid judgments if their observa-
Questionnaire: Validity data for adolescents and adults and its associa-
tions are quantified with psychometrically sophisticated assess- tion with non-suicidal self-injury. PLoS ONE, 10(12), e0145892. http://
ment tools and based on the criticisms related to self-reporting and dx.doi.org/10.1371/journal.pone.0145892
observer-rated measures for the assessment of mentalization. Pa- Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., & Plumb, I. (2001).
tients with egosyntonic problematics, in fact, may have issues in The “Reading the Mind in the Eyes” Test revised version: A study with
answering items that assess their self-reflexivity. At the same time, normal adults, and adults with Asperger syndrome or high-functioning
 10 GAGLIARDINI AND COLLI

autism. Journal of Child Psychology and Psychiatry, 42, 241–251. Fonagy, P., Luyten, P., Moulton-Perkins, A., Lee, Y. W., Warren, F.,
http://dx.doi.org/10.1111/1469-7610.00715 Howard, S., . . . Lowyck, B. (2016). Development an validation of a
Bateman, A. W., Bolton, R., & Fonagy, P. (2013). Antisocial personality self-report measure of mentalizing: The Reflective Functioning Ques-
disorder: A mentalizing framework. Journal of Lifelong Learning in tionnaire. PLoS ONE, 11(7), e0158678. http://dx.doi.org/10.1371/
Psychoanalysis, 11, 178 –186. http://dx.doi.org/10.1176/appi.focus.11.2 journal.pone.0158678
.178 Fonagy, P., Target, M., Steele, H., & Steele, M. (1998). Reflective func-
Bateman, A. W., & Fonagy, P. (2004). Psychotherapy for borderline tioning manual: Version 5. Unpublished manual.
personality disorder. Mentalization based treatment. Oxford, UK: Ox- Hausberg, M. C., Schulz, H., Piegler, T., Happach, C. G., Klöpper, M.,
ford University Press. Levke Brütt, A. L., . . . Andreas, S. (2012). Is a self-rated instrument
Bateman, A. W., & Fonagy, P. (2016). Mentalization-based treatment for appropriate to assess mentalization in patients with mental disorders?
personality disorders: A practical guide. Introducing Oxford Clinical Development and first validation of the Mentalization Questionnaire
Psychology. Advance online publication. http://dx.doi.org/10.1093/med: (MZQ). Psychotherapy Research, 22, 699 –709. http://dx.doi.org/10
psych/9780199680375.001.0001 .1080/10503307.2012.709325
Betan, E., Heim, A. K., Zittel Conklin, C., & Westen, D. (2005). Coun- Huprich, S. K., Bornstein, R. F., & Schmitt, T. A. (2011). Self-report
methodology is insufficient for improving the assessment and classifi-
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

tertransference phenomena and personality pathology in clinical prac-

cation of Axis II personality disorders. Journal of Personality Disorders,
This document is copyrighted by the American Psychological Association or one of its allied publishers.

tice: An empirical investigation. American Journal of Psychiatry, 162,

890 – 898. 25, 557–570. http://dx.doi.org/10.1521/pedi.2011.25.5.557
Blackshaw, A. J., Kinderman, P., Hare, D. J., & Hatton, C. (2001). Theory Meehan, K. B., Levy, K. N., Reynoso, J. S., Hill, L. L., & Clarkin, J. F.
of mind, causal attribution and paranoia in Asperger syndrome. Autism, (2009). Measuring reflective function with a multidimensional rating
5, 147–163. http://dx.doi.org/10.1177/1362361301005002005 scale: Comparison with scoring reflective function on the AAI. Journal
of the American Psychoanalytic Association, 57, 208 –213. http://dx.doi
Blagov, B., Bi, W., Shedler, J., & Westen, D. (2012). The Shedler-Westen
.org/10.1177/00030651090570011008
Assessment Procedure (SWAP): Evaluating psychometric questions
Morey, L. C. (2014). Borderline features are associated with inaccurate
about its reliability, validity, and fixed score distribution. Assessment,
trait self-estimations. Borderline Personality Disorder and Emotion
19, 370 –382.
Dysregulation, 1, 4. http://dx.doi.org/10.1186/2051-6673-1-4
Bo, S., Sharp, C., Fonagy, P., & Kongerslev, M. (2017). Hypermentalizing,
Oakley, B. F. M., Brewer, R., Bird, G., & Catmur, C. (2016). Theory of
attachment, and epistemic trust in adolescent BPD: Clinical illustrations.
mind is not theory of emotion: A cautionary note on the Reading the
Personality Disorders: Theory, Research, and Treatment, 8, 172–182.
Mind in the Eyes Test. Journal of Abnormal Psychology, 125, 818 – 823.
http://dx.doi.org/10.1037/per0000161
http://dx.doi.org/10.1037/abn0000182
Colli, A., & Gagliardini, G. (2015). Modes of Mentalization Scale. Un-
O’Connor, B P. (2000). SPSS and SAS programs for determining the
published manual, Department of Humanities, “Carlo Bo” University of
number of components using parallel analysis and Velicer’s MAP test.
Urbino, Italy. Behavior Research Methods, Instruments, & Computers, 32, 396 – 402.
Colli, A., Tanzilli, A., Dimaggio, G., & Lingiardi, V. (2014). Patient Petersen, R., Brakoulias, V., & Langdon, R. (2016). An experimental
personality and therapist response: An empirical investigation. American investigation of mentalization ability in borderline personality disorder.
Journal of Psychiatry, 171, 102–108. http://dx.doi.org/10.1176/appi.ajp Comprehensive Psychiatry, 64, 12–21. http://dx.doi.org/10.1016/j
.2013.13020224 .comppsych.2015.10.004
Davidson, K. M., Obonsawin, M. C., Seils, M., & Patience, L. (2003). Skårderud, F. (2007). Eating one’s words, part II: The embodied mind and
Patient and clinician agreement on personality using the SWAP-200. reflective function in anorexia nervosa – theory. European Eating Dis-
Journal of Personality Disorders, 17, 208 –218. http://dx.doi.org/10 orders Review, 15, 243–252.
.1521/pedi.17.3.208.22148 Stevens, J. (2002). Applied multivariate statistics for the social sciences
Dimitrijević, A., Hanak, N., Altaras Dimitrijević, A., & Marjanović, Z. J. (4th ed.). Mahwah, NJ: Lawrence Erlbaum Associates.
(2018). The Mentalization Scale (MentS): A self-report measure for the Taubner, S., Kessler, H., Buchheim, A., Kächele, H., & Staun, L. (2011).
assessment of mentalizing capacity. Journal of Personality Assessment, The role of mentalization in the psychoanalytic treatment of chronic
100, 268 –280. http://dx.doi.org/10.1080/00223891.2017.1310730 depression. Psychiatry: Interpersonal and Biological Processes, 74,
Ensink, K., Duval, J., Normandin, L., Sharp, C., & Fonagy, P. (2018). 49 –57. http://dx.doi.org/10.1521/psyc.2011.74.1.49
Measuring reflective functioning in adolescents: Relation to personality Westen, D., & Nakash, O. (2005). Attachment Questionnaire (AQ) manual.
disorders and psychological difficulties. Adolescent Psychiatry, 8, 5–20. Unpublished manual.
http://dx.doi.org/10.2174/2210676608666180208161619 Westen, D., & Weinberger, J. (2004). When clinical description becomes
Fonagy, P., Bateman, A. W., & Luyten, P. (2012). Introduction and statistical prediction. American Psychologist, 59, 595– 613. http://dx.doi
overview. In P. Fonagy & A. W. Bateman (Eds.), Handbook of mental- .org/10.1037/0003-066X.59.7.595
izing in mental health practice (pp. 3– 42). Arlington, VA: American Yaghmaie, F. (2009). Content validity and its estimation. Journal of
Psychiatric Publishing. Medical Education, 3, 25–27.