Content uploaded by Bettina Nielsen
Author content
All content in this area was uploaded by Bettina Nielsen on Dec 06, 2024
Content may be subject to copyright.
Implementing complex attachment-based interventions: Perspectives on the
circle of security-Virginia family intervention
Bettina Nielsen
a,*
, Hanne Weie Oddli
b
, Kari Slinning
c
, Filip Drozd
c
a
Department of Educational Science, University of South-Eastern Norway, Norway
b
Department of Psychology, University of Oslo, Norway
c
Department for Infant Mental Health, Regional Center for Child and Adolescent Mental Health, Eastern and Southern Norway, Norway
ARTICLE INFO
Keywords:
Attachment
Intervention
Circle of security
Parent-child relationships
Child safety
ABSTRACT
There are many promising attachment-based interventions that are accumulating an evidence-base, one of which
is Circle of Security (COS). The COS-Virginia Family (COS-VF) model is one of several versions of COS. The aim
of the study was to evaluate strengths and weaknesses of COS-VF from therapists’ viewpoint. Furthermore, we
wanted to examine opportunities to further develop the intervention, to make it more efcient for therapists and
families. There were 25 therapists in this study (5 males and 20 females). Participants took part in a semi-
structured interview and online questionnaire. Results indicated that therapists had positive perceptions about
the effectiveness of COS-VF, though the lack of effectiveness studies on the intervention was perceived as a
weakness. They praised the graphic illustrations designed for COS. Furthermore, therapists stated that the
assessment tools used for COS-VF gave in-depth information that enabled them to efciently get to the root of the
families’ problems. Therapists argued that teaching parents the COS-language facilitated communication and
alliance between therapists and parents. However, therapists also mentioned the danger of parents developing a
pseudo-language without internalizing the behavior. As such, it was considered a weakness that they only see the
children at the beginning and end of treatment, which also threatens their ability to assure children’s welfare.
Thus, they recommended including home-visits as part of COS-VF. There is a need for evaluating the effec-
tiveness of COS-VF and assessment tools that have previously not been validated, to assure families get the help
they need and ensure children safety.
1. Introduction
There is strong potential for attachment-based programs to improve
parental behavior and promote healthy development among vulnerable
children (Bakermans-Kranenburg et al., 2003; Letourneau et al., 2015;
Mountain et al., 2017; Wright et al., 2015). The number of attachment-
based interventions are increasing and accumulating an evidence base
that appears promising (Bakermans-Kranenburg et al., 2003; Letourneau
et al., 2015; Mountain et al., 2017; Wright et al., 2015). Attachment-
based interventions are believed to alleviate children’s distress, in-
crease parental sensitivity, and lead to a feeling of security in the
parent–child relationship (Blome et al., 2010). Furthermore, when
reviewing attachment-based interventions, Berlin et al. (2016) found
that the primary goals of the interventions was to target parental in-
ternal working models and parenting behaviors, promote parental
reection, and help parents make explicit connections between their
own histories and the parenting behaviors they do and do not replicate.
One attachment-based intervention that has received particular atten-
tion the last two decades is Circle of Security (COS; Marvin et al., 2002).
However, randomized controlled trials show mixed results in support of
the original therapeutic, group-based COS intervention (Dehghani et al.,
2014; Ramsauer et al., 2019).
1.1. The circle of security intervention
The Circle of Security (COS) intervention was developed by Marvin
and colleagues (2002) and delivered as group therapy. They argued that
behavioral and emotional problems in children could often be traced
back to attachment problems. Thus, parents could learn about strategies
to improve the attachment relationship. As a result, their children would
* Corresponding author.
E-mail addresses: Bettina.Nielsen@usn.no (B. Nielsen), h.b.w.oddli@psykologi.uio.no (H.W. Oddli), kari.slinning@r-bup.no (K. Slinning), lip.drozd@r-bup.no
(F. Drozd).
Contents lists available at ScienceDirect
Children and Youth Services Review
journal homepage: www.elsevier.com/locate/childyouth
https://doi.org/10.1016/j.childyouth.2024.108048
Received 15 December 2023; Received in revised form 7 November 2024; Accepted 20 November 2024
Children and Youth Services Review 169 (2025) 108048
Available online 28 November 2024
0190-7409/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (
http://creativecommons.org/licenses/by/4.0/ ).
be more likely to achieve healthy- development and attachments
throughout their lifespan; All of which was the basis for the intervention
(Powell et al., 2014).
The developers designed a graphic illustration (the Circle of Security:
See Fig. 1) of the most salient features of a secure attachment relation-
ship. The diagram simplies attachment theory for parents and is used to
help them decode their children’s behaviors as an expression of their
needs. Marvin and colleagues (2002) integrated this graphical illustra-
tion into an early treatment protocol which was a 20-session group
model where parents met weekly to review videos of themselves and
their children.
The aim of the intervention was to: 1) Enhance the parents awareness
and suitable reactions to the child’s cues when the child moves away to
explore or returns for comfort and reassurance, 2) improve their ca-
pacity to consider and understand both their own and the child’s be-
haviors, thoughts, and emotions related to their attachment and
caregiving interactions, 3) reect on past experiences that inuence
their current caregiving practices (Powell et al., 2015).
The core concepts of the intervention comprise Ainsworth’s and
colleagues (1978) ideas of a Secure Base and Safe Haven. Marvin et al.
(2002) argued that children are more likely to achieve secure attach-
ments if their parents act as a safe base from which they can explore
their environment and a secure haven where their children can return,
when faced with fear or threat.
The original COS group model has since spawned several adaptions
to make the intervention more accessible (Powell et al., 2014). Thus,
COS is no longer just one intervention, but rather a group of different
versions of the intervention. Furthermore, as the intervention became
increasingly popular, it was disseminated in several countries (e.g., USA,
Sweden, Australia, Norway, Japan, and Germany).
1.2. Circle of security-Virginia family model
One of the earliest adaptations of the original COS intervention, was
the individual treatment model (Circle of Security–Virginia family
[COS-VF] model). To date, there are no effect studies on COS-VF.
However, the structure of training and supervision of therapists and
the intervention itself, is, in many ways, similar to the original inter-
vention. The main difference is the group versus the individual treatment
format. COS-VF is a therapeutic model designed for severe cases, where
children (11 months – 6 years) are considered at great risk of developing
or have already developed attachment problems (for inclusion and
exclusion criteria see Table 1). The aim of the intervention is to promote
healthy parent–child relationships (i.e., secure attachment) by means of
increased parental sensitivity and reective functioning. Each family
receives psychotherapy over a period of about 20–30 weeks, depending
on the severity of their problems. Furthermore, COS-VF is a manual-
based invention with six phases (Marvin & Whelan, 2006; see Table 2).
Fig. 1. The Circle of Security illustration.
Table 1
Inclusion and exclusion criteria COS-VF (Marvin & Whelan, 2006).
Inclusion criteria: Exclusion criteria:
1. Parent-child dyads with a relationship
characterized by behavioral- and
emotional difculties.
2. Children who have experienced
trauma or ruptures in the relationship
with their caregivers.
3. Children with behavioral and
emotional difculties
4. Parents should have abilities within a
minimum of 2 of the following areas: (a)
curiosity about children and their inner
life, (b) an ability to see things from the
child’s perspective, (c) an ability to
perceive the child in a developmental
perspective, (d) an ability for reection
about own thoughts, emotions, and
behavior, specically towards their
child, (e) willingness to be counseled by
the therapists.
1. Children with severe behavioral
problems and problems with acting out.
2. Children with autism.3. It is also
considered unlikely that parents will be
able to take full advantage of the
program if they are: (a) in a severe
crisis, (b)
experiencing a problematic divorce or
custody battle, (c) actively abuse drugs
and/or alcohol, (d) physically violent,
(e) have moderate to severe cognitive
impairments, (f) have a severe
psychiatric illness.
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
2
1.3. Central terms used in COS-VF
As part of the intervention’s psychoeducation, parents are intro-
duced to the COS-VF framework and introduced to what is often referred
to as “the COS language”. That is, they learn different terms and phrases
therapists use when talking about attachment and different aspects of
parent–child behaviors and interactions. To better understand the
nding in this study, some central terms in COS-VF will be presented
(Marvin, 2006; Marvin & Whelan, 2006; Powell et al., 2014).
1.3.1. Cues and Miscues
A person’s direct or indirect effort to make his or her needs known to
another person, are referred to as “cuing”. However, if someone covers
their needs, this is referred to as “miscuing”. Miscues are used to conceal
a need that the individual has learned is unsafe to express openly. Thus,
they can include behaviors such as looking- or walking away, even if a
person needs comfort and/or support. Therefore, parents need help to
decode their children’s miscues and to embrace the children’s needs.
1.3.2. Shark Music
Shark music is essentially what others refer to as “ghosts in the
nursery” (Fraiberg et al., 1975) or “voices from the past” (Dozier et al.,
2008). It represents the strong emotional responses that children’s
behavior can occasionally trigger in parents. These emotions are often
caused by the parent’s procedural memories of previous attachment
experiences. Shark music is usually unconscious and can prevent parents
from seeing what needs are reected in their child’s behavior. Thus,
COS-VF aims to help parents explore and label these emotional re-
actions, as well as teach them how to prevent their shark music from
blurring their perceptions of their child’s needs (Marvin & Whelan,
2006; Powell et al., 2014).
1.3.3. Being-With
All children have an innate need to “Be With” a parent who can
provide protection, comfort, and necessary social interaction (Powell
et al., 2014). The main aspect of developing a secure attachment is that
the child can learn to trust their parent to provide physical protection
and be emotionally available when needed (Marvin & Whelan, 2006;
Powell et al., 2014). The caregiver’s ability to be present with the child
forms the basis for the child’s emotional and social development. By
“Being-With” the child, the parent provides responsive caregiving and
thus, has the greatest hope of meeting the child’s needs. The interven-
tion helps parents understand their children’s attachment needs and
tries to teach them how to meet those needs. This because, children’s
knowledge that their caregiver is emotionally available to Be-With them
when needed, is crucial for the development of secure attachment.
1.3.4. Core Sensitivities
Most people have unconscious and unresolved fears in attachment
relationships, and it is not uncommon for children to become sensitized
to certain parent struggles that COS developers refer to as “core sensi-
tivities” (Powell et al., 2014). Each sensitivity represents an internal
working model that guide the relationship to avoid unregulated affect.
The intervention is intended to help parents reect upon how their
sensitivities affect their interactions with their child.
1.4. COS-VF treatment assessments
The original COS model uses the Strange Situation Procedure (SSP:
Ainsworth et al., 1978) and follow the Cassidy and Marvin (1992) pre-
school system assess attachment patterns of parent–child dyads (Powell
et al., 2014). However, in COS-VF the Secure Base Safe Haven coding
system (Marvin & Whelan, 2007) is used to classify attachment
behavior. Furthermore, some adaptions have been made to the SSP. For
instance, in the original SSP, dyads are left in the room for 3 min at the
start of the procedure, while in COS-VF they are left for 5 to 10 min. This
is to allow more “free play” moments. In COS-VF these moments are
lmed to make video material for the intervention. The COS-VF pro-
cedure is for children aged 4 ½ −6, as opposed to the original version
where previous research has included children from approximately 1–7
years (e.g., Marvin et al., 2003; Huber et al., 2015). Furthermore, COS-
VF developers had experienced that, children in this age group (i.e., 4 ½
−6 years), became comfortable with the stranger, which consequently,
made their attachment behavior very subtle and hard to classify. Thus,
in COS-VF the stranger does not return after the last separation.
During the SSP, it is central to look for the linchpin struggle (Powell
et al., 2014). That is, therapists try to identify the critical and recurring
challenge or difculty that the parents face in understanding and
responding to their child’s emotional needs. This struggle is seen as the
core issue that, if addressed, can lead to signicant improvements in the
parent–child relationship, fostering secure attachment. Video recording
from the SSP is used in sessions to allow parents to observe the in-
teractions with their child and discuss these interactions with their
therapist. The aim is to observe where, on the circle (see Fig. 1 above),
they struggle and where they succeed, when dealing with their child’s
needs and difcult emotions; The core of the intervention (Powell et al.,
2014). The treatment manual also suggests evaluating parents after
treatment using the SSP.
Therapists use the COS-interview (COS-I), designed to help identify
the interplay among core beliefs, feelings, behaviors, and perceptions.
Furthermore, COS-I seeks to uncover parent’s capacities as well as the
nature and meaning of their defensive capacities. COS-I was created by
the COS developers; however, it is based on exiting state of mind in-
terviews (Andrews & Coyne, 2018), such as the Adult Attachment
Interview (George et al., 1996), the Working Model of the Child Inter-
view (Zeanah et al., 1994) and the Parent Development Interview (Aber
et al., 1985). COS-I is divided into three sections where parents are asked
reective questions concerning: 1) their understanding of their child’s
SSP behavior, 2) their daily life with their child, 3) their own life,
growing up with their primary caregiver.
After these assessments, therapists use the information to customize
an individualized treatment plan for each family (Powell et al., 2014). The
manual also recommends using other assessment tools, such as the
Table 2
The six phases in the COS-VF treatment manual.
Phase Objective Description
1 Assessment Families are assessed using the Strange
Situation Procedure to evaluate the quality of
the attachment pattern and the COS interview
for insight into parent’s own attachment
history. Information from the assessments is
used to develop a treatment plan.
2 Establish a secure base
and safe haven
Therapists aim to establish a supportive and
safe environment for parents by using videos of
other parent–child relationships and exercises
dened by the manual.
3 Learning the COS
framework
Parents are provided with an introduction to
attachment theory using the COS graphics.
4 Develop observational
abilities
Parents learn to enhance/develop observational
abilities by use of videos of other families,
focusing on identifying what needs these
children are demonstrating, and using the COS
graphics as a roadmap.
5 Increasing reective
functioning
Parents gain increased awareness about the
strengths and weaknesses of their own
relationship with their child and explore how
previous attachment experiences can affect
their current attachment patterns.
6 Empathic shift,
assessment of change
The new knowledge is integrated as parents
gradually receive automatic access to their
internal working models and the associated
COS vocabulary. The Strange Situation
Procedure can be used to evaluate treatment
effects.
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
3
Parent Stress Index-Short form (Abidin, 1995) or the Symptom Check
List 25 (Derogatis et al., 1974).
1.5. Challenges with Attachment-Based interventions
While professionals see the advantage of attachment theory, the
process of transmitting these ideas into practice are often vague and less
tangible (Blome et al., 2010). Thus, there are many challenges in
developing and implementing attachment-based interventions. For
example, they are complex to implement properly and may require
much effort from the therapists. Furthermore, the structure of mental
health and social welfare services is often segmented, instrumental, and
short-term, which can challenge the ability of clinicians to relate to
families and commit to a planned intervention protocol (Blome et al.,
2010; Nielsen et al., 2020). Furthermore, many programs and in-
terventions have a disconnect between the design context and the
implementation context, which can lead to decisions regarding the
design that are not applicable to the constraints of the context in which
they will be used. For example, they can be too expensive, impractical,
or even impossible to use under real-life constrains; a problem that has
often been overlooked in mental health services (Lyon & Koerner, 2016).
Previous research show that there are challenges in introducing and
using, both COS-VF and the original group model, in mental health and
social services (Hoffman et al., 2006; Nielsen et al., 2020). One study
also indicated that even with the supposedly “simplest” of the COS
models (i.e., COS-P), it can be difcult to use the intervention in its
entirety as intended (Maupin et al., 2017). Currently, there are few
studies, with some notable exceptions such as Maupin and colleagues
(2017), that have systematically investigated aspects of attachment-
based interventions, which may have important consequences for their
uptake, use, and implementation in practice.
It has been argued that Roger’s Diffusion of Innovation theory (2002)
can be useful to inform the understanding of the process adoption and
sustainability of innovations in child welfare practices (Blome et al.,
2010). According to this theory there are ve components that deter-
mine the rate of adoption of an intervention: 1) relative advantage (i.e.,
the extent to which an innovation is perceived as an improvement to
current practices), 2) compatibility (i.e., the degree to which the inno-
vation is viewed to be consistent with existing values and needs of
possible users), 3) complexity (i.e., how difcult the innovation is to
understand and use), 4) trialability (i.e., how much experimentation can
be done with the innovation), and 5) observability (i.e., how visible the
results will be to others). Thus, the innovation will be more promptly
adopted if the alterations are more exible in its implementation, out-
comes are more visible to others, and the innovation is perceived to be
well-matched with existing values of users. On the other hand,
complexity is negatively related; the more complex the innovation, the
slower the innovation will be adopted. Based on this theory the
complexity inherent in attachment-based interventions can challenge
implementation. However, since the values of attachment theory are
consistent with that of welfare services, this may still resonate with
workers and managers (Blome et al., 2010). Nonetheless, it is important
to identify which aspect of these interventions’ therapists deem impor-
tant and essential for implementation, which they perceive as short-
comings, and to identify potential areas for further adaptations and
development.
1.6. Aims of this study
The aim of this study is to assess the strengths and weaknesses of
COS-VF from therapists’ perspective. Furthermore, the intention is to
determine whether there are possible opportunities to adapt and further
develop the COS-VF model in a way that makes it more efcient for both
therapist’s and families in practice. Thus, this study can contribute to
facilitate the implementation of the model and improve COS-VF for
therapists and families. Moreover, ndings from this study can be
generalizable and utilized for other attachment-based interventions,
video-based treatments, and interventions in mental health and welfare
services in general.
2. Method
2.1. Design
This study used a mixed-methods approach with a one-phased
triangulation design, in which the qualitative data were the most
dominant source of information (Creswell, Plano-Clark, & Garrett,
2008). The study was approved by the Norwegian Centre for Research
Data.
2.2. Participants
At the time of the study, there were 54 certied COS-VF therapists in
Norway, all of whom were invited to participate. The researchers were
given access to their contact information via the COS-VF program pro-
viders (The Regional Center for Child and Adolescent Mental Health,
Eastern and Southern Norway) and participants were contacted via
telephone or Therapists who wanted to participate or learn more about
the study, received an email with a link to the study information and
consent form. They could participate by 1) responding to an online
questionnaire and/or 2) taking part in a semi-structured interview. In
total, 25 therapists (46 % of those invited) participated in the study. Of
these, 24 therapists (96 %; 5 males, 19 females) completed the online
questionnaire, while interviews were conducted with 19 therapists (76
%; 4 males, 15 females). One interview was excluded due to low-quality
audio. Additionally, one participant took part only in the interview
without completing the questionnaire. Three of the COS-VF therapists in
the sample were also COS-VF supervisors.
2.3. Procedure
After signing consent forms, participants could answer the online
questionnaire immediately. Those who wanted to participate in in-
terviews, were contacted via telephone. All interviews were conducted
by the rst author (BN), either face-to-face, via Skype or telephone. This
was contingent on whether the participant lived near the researcher
and/or their personal preference. Three participants were interviewed
face-to-face, one via Skype, and 15 via telephone. All interviews were
audio recorded and transcribed prior to analysis. The duration of the
interviews ranged from 19 to 90 min.
2.4. Data
2.4.1. Quantitative
Therapists lled in an online questionnaire, including questions
about personal characteristics and delity to the treatment manual.
Items were scored on a 5-point Likert-type scale (e.g., 1 =“I do not
agree” to 5 =“I totally agree”, 1=“Never” to 5=“Always”) and pro-
vided an overview of which intervention components were actively used
by therapists (or not) to better assess the opportunities for further
development of the COS-VF method.
2.4.2. Qualitative
The semi-structured interviews included open-ended questions con-
cerning therapists’ perceived Strengths, Weaknesses, Opportunities, and
Threats (SWOT; Helms & Nixon, 2010) of the COS-VF intervention. This
interview method enables spontaneous descriptions that are minimally
inuenced by the researcher’s pre-conceptions. Furthermore, it does not
impose any assumptions nor specify any types of answers from in-
terviewees. The interview contains four sections, where each section
relates to each of the SWOT dimensions (see Table 3). Participants were
also asked probing questions throughout the interview.
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
4
The respective elements of the SWOT acronym are dened as fol-
lows: (S) a resource or a unique approach that facilitates the achieve-
ment of treatment goals, (W) a limitation, fault, or defect that can hinder
obtainment of treatment goals, (O) internal or external forces in the
work environment that may facilitate a more effective way of obtaining
treatment goals, and (T) any unfavorable situation in the work envi-
ronment that can be viewed as a barrier or constraint that limits the
obtainment of treatment goal.
2.5. Data analysis
Total numbers and percentages were calculated for therapist char-
acteristics while means and standard deviations were calculated for -
delity questions. The rst author conducted all interviews,
transcriptions, and edited the data. All identifying information was
removed from the transcript before the analysis, to ensure condenti-
ality. Data were then analyzed using the Framework method (Gale et al.,
2013) which involves seven steps: 1) transcription of audio -recorded
data, 2) familiarization with the data, 3) creating codes that represent
key concepts or themes, 4) developing a working analytical framework,
5) applying the analytical framework, 6) charting data into the frame-
work matrix, 7) interpreting the data.
Codes were data-driven and developed during the analysis (Braun &
Clarke, 2006). The rst author coded all the data, while the second and
last author read three interviews each and reported on any differences of
opinion on coding. Once all data material had been coded, codes were
indexed into categories related to whether they were perceived as
strengths, weaknesses, threats, or opportunities. The format of the
SWOT interviews should in theory have sorted the codes by the order of
the questions asked (i.e., “Tell me about the strengths/weaknesses/
threats/opportunities of the COS-VF intervention”); however, many
participants described threats when asked about strengths, weaknesses
and/or opportunity and vice versa, which made this sorting process
necessary to attain a framework that gave a better overview of the data.
The framework matrix feature in Nvivo v12 was used to create matrices
for each SWOT category.
The rst author made summaries for each individual case and topic
within the matrices, after which separate summaries were made for each
category. The last authors consulted throughout this process of the
analysis. Lastly, all authors reviewed the summaries individually, after
which research meetings took place were the different interpretations of
the data was discussed, and specic signicant themes were determined
as a group.
3. Results
3.1. Participant characteristics and delity questions
This study was part of a larger project on COS-VF, using the same
sample; thus, the information collected concerning therapists’ charac-
teristics is the same as previously reported in Nielsen et al., (2020).
Results showed that 80 percent of therapists worked full-time, most had
several years of experience working with COS-VF, and many partici-
pated in the COS-VF training within the rst years of implementation
(see Table 4). Therapists followed the inclusion and exclusion criteria
specied in the treatment manual to a reasonable extent, although it was
more common for therapists to use COS-I than any other assessment tool
Table 3
Interview questions.
Part 1: Open-ended SWOT-questions
1. Please tell me about the strengths you perceive regarding COS-VF
2. Please tell me about the weaknesses you perceive regarding COS-VF
3. Please tell me about the opportunities you perceive for improving the quality of COS-VF
4. Please tell me about the threats you perceive regarding COS-VF
Table 4
Participant characteristics (N =25).
Characteristics Total
sample
(n =25)
a
Interview
participants
(n =19)
a
Gender n (%)
Female 20.0
(80.0)
15.0 (79.0)
Male 5.0 (20.0) 4.0 (21.1)
Educational background n (%)
Social worker 10.0
(40.0)
5.0 (26.3)
Psychologist 8.0 (33.3) 6.0 (31.6)
Educator (i.e., pedagogue) 3.0 (12.5) 3.0 (15.8)
Other 3.0 (12.5) 3.0 (15.8)
Informant n (%)
Therapist 22.0
(88.0)
16.0 (84.2)
Supervisor 3.0 (12.0) 3.0 (15.8)
Years in current position (M ±SD) 5.2 ±2.8 5.5 ±2.8
Work situation n (%)
Part-time 4.0 (16.7) 4.0 (21.0)
Full-time 20.0
(83.3)
15.0 (79.0)
Years since start of COS-VF education (M ±SD) 7.5 ±1.8 7.6 ±1.7
Years working with COS-VF (M ±SD) 5.2 ±1.8 5.2 ±1.8
Number of colleagues working with COS-VF (M
±SD)
1.8 ±1.5 1.9 ±1.7
Current organization n (%)
Child and adolescent psychiatric clinic 4.0 (16.7) 2.0 (10.5)
Family center 3.0 (12.5) 3.0 (15.8)
Privat practice 3.0 (12.5) 2.0 (10.5)
State child welfare service 2.0 (8.3) 0.0 (0.0)
Municipal child welfare service 5.0 (20.8) 4.0 (21.0)
Other 7.0 (29.2) 6.0 (31.6)
a
Numbers and percentages adjusted for missing values (i.e., counted actual
number of observations and used these in the denominator for percentages
rather than total sample size).
Table 5
Means and standard deviations for treatment delity questions (N =24).
Inclusion criteria in the COS-VF manual are followed as prescribed 3.6
(0.6)
Exclusion criteria in the COS-VF manual are followed as prescribed 3.4
(0.7)
Which of the following tools do you use for the initial assessment?
The Strange Situation Procedure (SSP) 3.5
(0.7)
The COS–Interview 3.8
(0.5)
Parent Stress Index–Short Form 0.7
(1.0)
Symptom Check List 25 or other validated measures of anxiety and
depression
0.5
(0.7)
Do you develop an individual treatment plan for each family? 3.1
(0.9)
Do you introduce the term miscues to the parents? 3.9
(0.3)
Do you introduce the term shark music to the parents? 4.0
(0.2)
Do you ask the parents to bring circle stories to therapy sessions? 3.8
(0.6)
Do you use the SSP to evaluate families at the end of COS–VF 2.5
(1.3)
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
5
(see Table 5). The results also show that most therapists introduced the
concepts of “shark music” and “miscues” and requested that parents
bring ‘circle stories’ to treatment sessions. Furthermore, assessing
dyadic interactions at the end of treatment using the SSP was less
common than at the beginning, which was expected since the treatment
manual does not specify that dyadic interactions must be assessed after
treatment.
3.2. Analysis of interview data
When analyzing the interview data, 324 codes were identied (e.g.,
“cues and miscues”, “shark music”, and “delity to the COS-VF
method”), which generated 6 themes.
3.2.1. Theme 1: Providing Comprehensive and Accessible Psychoeducation
Many therapists emphasized that COS developers have managed to
make the complicated attachment theory comprehensible and accessible
in a new way, which has opened the door to many new possibilities. This
included the COS language, the graphics, and the SSP videos. As stated
by one therapist: “the visualization of the theory is probably the in-
terventions’ greatest strength. Most people, parents, and therapists nd
visual tools helpful. It creates something we have in common, something
to refer to”. Therapists also discussed how circle illustrations were very
helpful when trying to teach parents to focus on their child’s needs
instead of their behavior.
The COS Language. Several therapists mentioned the advantages
gained by teaching parents the COS language. For instance, one thera-
pist said that by teaching parents to talk about attachment behaviors
using the COS language, they were able to “talk to each other in a way
that makes the “expert label” therapists often get, disappear.” It was also
mentioned how this made it possible to speak to each other “on the same
level”. Furthermore, many described that teaching parents about shark
music, cues, and miscues, was very useful. One therapist had noticed
various parents talking to each other after receiving COS-VF treatment
and using phrases like “your shark music was probably turned on by
that” or “that day I was at the bottom of the circle”. This was benecial
because they immediately started focusing on the relationship, instead
of the person.
Shark Music. Therapists praised the shark music component of COS-
VF, and felt it was an important part of the intervention. Furthermore,
therapists argued that it can be very redemptive to focus on how the
parents themselves were met by their parents and it makes it easier to
work on parents’ challenges when they have reected and explored their
shark music. By calling it shark music, it becomes something normal,
instead of something scary; a natural part of everyone’s life. However, it
was also stated that shark music is not necessarily a concept parents
understand right away. Moreover, it can be hard for therapists to address
the parent’s greatest struggles. Someone’s shark music can involve
difcult things from their past, and events they have been exposed to,
that reappears in their consciousness. One therapist revealed that it
could be hard to talk to parents about “what from your past could be
playing tricks on you today?”. Moreover, it could be hard to uncover
what it was. Additionally, therapists explained that shark music could
touch upon something very emotional and dealing with it could cause a
lot of anxiety for the parent. At times, to the extent that they would have
to take a break from treatment. Other times they might have to end
treatment, to enable parents to receive treatment for their personal
problems.
Video-Materials. Most therapists were positive about using videos
of other families in their sessions to teach parents about attachment and
how to reect on parent–child interactions. This allowed parents to
learn by watching others, before they observed their own video from the
SSP.
Most considered the videos used in COS-VF to be old and of poor
quality. However, one therapist stated that “It doesn’t really matter if
you are able to hear them [dyads in the video] properly or not… The
more you have to explain to those in front of you [parents] the better…
you are forced to be more thorough”. Furthermore, it was more
important to evaluate what the parents were ready and/or able to un-
derstand, “If you make the wrong judgment about that, then it doesn’t
matter what kind of movie you show them or what the quality is.”.
Some therapists wanted more variation in the video materials, as this
would make it easier to customize the treatment better to individual
families with different problems. This would also make it easier for
parents to identify with the video-materials. Others felt some of the
videos were too “American”, which also made the interactions feel less
familiar for Norwegian parents. Thus, they wanted Norwegian material.
Albeit there were some therapists who mentioned the challenges of
obtaining materials given the country’s small population size. They
discussed how this made it more likely that families in the video-
materials could be recognized. Furthermore, even if parents agreed to
share videos from the SSP, it would still be challenging to protect the
rights and condentiality of the child.
3.2.2. Theme 2: Obtaining Reliable Assessments of the Families
Therapists were positive about the assessment tools used in COS-VF.
These tools gave them considerable in-depth information about the
family relationships, and the parent’s strengths and weaknesses at once.
Furthermore, the tools allowed them to quickly get to the root of the
family’s problems. One therapist felt surprised to nd that the struggles
identied at the start of the intervention corresponded well with the
struggles observed when working with the families over time: “Previ-
ously I would have had to spend a lot more time before I gured these
things out”.
Some therapists mentioned that they had not learned how to assess
reective functioning during their COS-VF training, even though was an
important factor to assess. Therefore, there was a desire for assessment
tools that could more efciently and easily measure parent’s reective
capacities, to make sure they were able to benet from treatment.
The SSP. Therapists were generally positive about using the SSP and
saw it as benecial to use videos of the parent–child dyads as part of
therapy sessions. Nevertheless, some mentioned that it was a challenge
that therapists could not conduct the SSP alone as it is necessary for
someone to act as “the stranger”. Many also found it difcult to evaluate
the SSP by themselves. They worried they might miss details and
consequently offset the suitable direction of the treatment by focusing
on the “wrong” linchpin struggle. Some therapists, who worked alone
with COS-VF, would prefer a supervisor or someone to discuss cases with
and get advice. One supervisor argued that to get reliable assessments of
the SSP, COS-VF therapists “could make a system where the different
therapists scored each other’s videos”. However, this might also be
problematic because: “… at a workplace people are so involved in each
other’s cases. Hence, ideally, they should get a therapist that worked
elsewhere to score it”. The supervisor stated further that, though this
would be preferable, it seemed unrealistic.
3.2.3. Theme 3: Ensuring Child Safety
Several therapists considered it a weakness that they were only
supposed to see the children in the beginning and end of treatment. This
made it hard to ensure the children were safe at home. Even though
parents were asked to bring ‘circle stories’ to therapy sessions, it was
difcult to be sure the stories were reliable. Parents could attempt to be
perceived in favorable ways or as explained by one therapist: “They
might think they are behaving in a secure manner, even when they are
not”. It was argued that it was difcult to be sure if parents had really
understood everything and not just saying what the therapists wanted to
hear. In other words, there was a risk of parents developing a pseudo-
language, implying that they may seem to have understood an issue,
while, as expressed by one therapist, “they have not been able to inte-
grate it as a natural response in relation to their child”. Thus, it was
difcult to know if the knowledge was truly integrated or just pseudo-
mentalization.
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
6
Several therapists argued that using home-videos would ensure that
therapists and parents observed the same parent–child interactions;
thus, therapists would not have to worry if parents’ circle stories were
true, decorated realities, and/or ction. Therapists also maintained that
this could be of help for those who have less reective abilities and nd
it difcult to tell circle stories in therapy sessions. Moreover, it would
provide therapists with additional observations of the children and
assure their well-being. Some had, in fact, already implemented this as a
part of the intervention.
Some therapists mentioned that because of the length of the inter-
vention, it can take time before they start using the SSP-videos of the
dyads in therapy sessions. Consequently, videos can become outdated,
as children develop rather quickly. One therapist had experienced par-
ents who claimed, “that’s how it was earlier, it’s not like that anymore”.
Which was yet another argument for including more updated video
material throughout the intervention.
Certain therapists felt it was a weakness with COS-VF that they had
to spend so much time teaching parents about attachment patterns and
watching videos of other families; especially if they were worried about
the children’s welfare. As expressed by one therapist: “One of the dis-
advantages with the family model is that it takes quite a long time before
you start working with the video of the dyad. You spend weeks watching
videos of other people. And if you are very worried, then it can take too
long before you start the actual work”. One supervisor also mentioned
this issue; however, this therapist further emphasized that “when you
are a bit more experienced, you are more exible; thus, you skip some of
it [the psychoeducation] and start watching lms of the family much
faster”. Some therapists pointed out that because the model is compre-
hensive, it is not appropriate for severe cases where there is a need for
immediate change to ensure the children’s safety.
3.2.4. Theme 4: Highlighting the Therapists’ Role and Inuence on
Treatment Outcomes
Therapists emphasized the importance of creating a secure envi-
ronment (i.e., a safe haven): “I need to be aware of my role as the one
who must create enough security for parents to dare see their own
vulnerability, and that it is acknowledged … I can acknowledge the
parent’s vulnerability and organize their feelings.” To this end, many
found it helpful to learn about different parental core sensitivities.
Therapists claimed this taught them how to deal with different parents
whom they might nd challenging to work with. However, they also
discussed how their own sensitivities “affects how we are as therapist,
when we meet different people”. Thus, some wished they had learned
even more about sensitivities: “It is wise to be aware of what it is about
some people that makes you experience them as more difcult and
challenging than others, who you feel empathy with and who you don’t,
and how you can work on that. Work on yourself as a counselor or
therapist”.
Clinical Experience. There were several accounts on how experi-
ence plays an important role for successfully using the intervention.
Therapists discussed that you must know the manual well, to be
completely and mentally “present” in therapy sessions: “You cannot be
‘in your head’ looking for what you should say according to the
manual… You use it as a point of reference, but exibly.” Furthermore,
the therapist stated that it was possible to customize the intervention to
the specic needs of each individual family because the treatment
manual “is clear about where you’re supposed to start and where you’re
supposed to end up”. One therapist explained that sometimes you
encounter dilemmas during the intervention, where you must rely on
your own clinical judgement and therapeutic skills. For example, when
evaluating how much time you should spend focusing on the parent’s
shark music: “Sometimes I am not sure how much I’m supposed to “rub
it in” by showing bad moments from the strange situation procedure, to
avoid making the parents feel too down”. Thus, clinical judgement and
therapeutic skills can determine how well therapists are able to work
through the different phases of the manual.
The COS-VF Treatment Manual. Several mentioned that the
treatment manual lacked clarity. As expressed by one therapist, “It
would be better if the manual was more concrete, because it would be
easier to let go of it, without feeling insecure”. Furthermore, it was
proposed that it would be easier if there was a clearer evaluation about
what you are expected to achieve throughout the different phases of the
manual. Some argued that the manual for the Circle of Security-Virginia
group model (COS-VG) was better in that regard. A supervisor provided
an alternative perspective and argued that COS-VG is even more
manual-based in the way that you follow a certain treatment pathway,
while COS-VF demands more therapeutic experience and skills.
According to the manual, therapists are not supposed to provide
parents with too much information about what is going to happen
throughout the intervention. Therapists discussed that if they did, the
intervention would not work: “It would probably not make sense and it
could become overwhelming”. One therapist, however, expressed that
sometimes it could make you feel like you were misleading parents to get
involved in something they might not want to do. The therapist ques-
tioned if “it would be possible to be more open with parents and give
them a reasonable explanation, without scaring them to death”. Another
therapist also mentioned that they were not supposed to tell parents
about the results of the assessments evaluation (i.e., of family) until
quite late in the intervention, which was regarded as a weakness: “I nd
the counseling easier if we have told them [parents] something about
what their struggles are … Our experience is that it is useful to give the
parents some feedback early on”.
3.2.5. Theme 5: Dening the Target Population
The general opinion was that COS-VF is not appropriate for
everyone, although there were different thoughts of whom the inter-
vention was suitable for. For example, one supervisor argued that “it is
important to follow the inclusion and exclusion criteria for the inter-
vention to work”, while others felt that they had to assess each case
individually. Some believed that in certain cases, the intervention could
also be helpful for families who were ineligible according to the criteria,
for example older children (i.e., >6 years).
A few therapists felt that the way COS-VF was designed, made it
suitable for parents with low reective abilities, while others stressed
that it was important that parents had a certain level of reective
functioning to benet from the intervention. Clearly, there were dif-
ferences of opinion. Thus, a statement from one of the supervisors could
be a necessary next step for further investigation. The therapist in
question, argued that “it is important to determine more clearly who can
benet from the intervention and who cannot”. Furthermore, if they
wanted to include families within the target population of the inter-
vention (i.e., that are currently regarded as ineligible according to the
inclusion/exclusion criteria), it was important to assess what changes
were needed for the intervention to allow these families to benet
optimally from treatment.
3.2.6. Theme 6: Uniting Positive and Realistic Expectations of Treatment
Outcomes
Many were very positive about COS-VF and one supervisor even
mentioned that “COS has received a status as a miracle cure”. Conse-
quently, many praised the effectiveness of the model: “I have never
worked with something that has greater potential for changing insecure
attachment patterns” and “this [COS-VF] is the best we’ve got”.
Furthermore, one therapist stated: “Because it [COS-VF] focuses on the
relationship, instead of the personality or personality structures, parents
can free themselves from shame and guilt and start thinking that they
have an opportunity to change…”.
Therapists argued that COS-VF is compatible with what is known
about children’s development and behavior in many different disci-
plines such as attachment, developmental psychology, trauma, and
neuropsychology. They said there was much condence in the model
and that, in general, people were not skeptical of COS-VF. Additionally,
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
7
they expressed that parents also had positive experiences with COS-VF.
For instance, one therapist stated “COS makes sense. I don’t think I’ve
ever experienced working with parents who thought it was just
nonsense”. But even so, therapists mentioned that the lack of studies
evaluating the effectiveness of the intervention was a weakness, which
made it hard to convince some service leaders why they should put effort
into implementing the intervention.
Even with all the praise that COS-VF received, a therapist asserted
that it is important to understand that “one method by itself, cannot
work wonders”. Several therapists agreed that you need to be realistic
about what you can accomplish with the intervention, and that the
extent of the family’s problems can affect how much you can achieve.
Families can improve but may not necessarily change from a disorga-
nized to a secure attachment pattern.
4. Discussion
In this study, the strengths, and weaknesses of COS-VF from thera-
pist’s perspective was assessed and the opportunities for further devel-
opment of the intervention was explored. However, the ndings can also
be utilized for many interventions in mental health and social services as
COS-VF has many similarities with other COS-, attachment-based-,
parent–child-, and video-based interventions. The qualitative analysis
identied six themes: 1) Providing comprehensive and accessible psy-
choeducation, 2) Obtaining reliable assessments of the families 3)
Ensuring child safety 4) Highlighting the therapists’ role and inuence
on treatment outcomes, 5) Dening the target population, 6) Uniting
positive and realistic expectation of treatment outcomes. The qualitative
ndings were further enriched by quantitative data from the online
questionnaire, both of which will be discussed in the following section.
The graphic COS illustrations were highly appreciated. Therapists
emphasized that COS developers managed to make complicated
attachment theory comprehensible and accessible. Marvin and col-
leagues (2002) described that their goal was to present Ainsworth and
colleagues’ (1978) ideas in a “user-friendly” way that made it cogni-
tively and emotionally accessible to parents. Our results indicate that
they have, indeed, successfully achieved this goal and that the graphics
are one of the major strengths of COS-VF, and most likely the other COS
interventions.
Another strength was the “COS language” which removed the
“expert” label and allowed therapists and parents to communicate on the
same level. Giving parents a language that can describe difcult feelings
in an easy way, may help them work through their problems more
successfully. This is important considering that putting feelings into
words may be one of the best ways of managing negative emotional
experiences (Lieberman et al., 2007). Nevertheless, therapists stressed
the threat of pseudo-mentalization, and were reluctant to rely solely on
parent’s circle stories. Knowing how to use the COS-language to describe
circle stories, does not mean parents have managed to translate the COS-
language into natural responses in their daily behavior. This could just
mean that they learn how to describe what they think their therapists
“want to hear”. Thus, therapists maintained that including more meet-
ings with children would be a better way of ensuring the children’s
continued improvement, as well as their safety. They also argued that
videos from the SSP can become outdated, and some would prefer more
updated video-materials of the families throughout the intervention.
However, even though many discussed these issues as a weakness with
COS-VF, the questionnaire data revealed that therapist do not always
choose to meet the children at the start or end of treatment for
evaluation.
Many therapists praised COS-VF and seemed satised with treatment
outcomes. However, the fact that therapists do not always perform pre-
and post-treatment evaluations and, thus, rely on their own and the
parents’ perceived usefulness of treatment, may have disadvantages.
This is because perceived usefulness does not necessarily mean the
treatment was effective. On the one hand, therapists may, for example,
be susceptible to “self-assessment bias”, which can lead them to think
patients have improved more than they have (Walsh et al., 2012). On
the other hand, parents in, for example, child protection services may be
susceptible to self-serving (Bernstein et al., 1979) or social desirability
biases (Crowne & Marlowe, 1960). Consequently, an argument can be
made for routine progress and outcome monitoring, using formal mea-
surement methods, including blind observations of videos to reduce bias
of therapists and parents (Tracey et al., 2014; Walsh et al., 2012). In
COS-VF, this may include observations of parent–child interaction from
home-visits or at the therapist’s ofce, and/or feedback on video-based
circle stories. As argued by Bick and Dozier (2013); by following
parent–child interactions across longer periods of time, they were more
likely to observe parent’s ability to adjust their behavior to the chil-
dren’s needs. Thus, including more dyadic observations throughout the
intervention, could generate some advantages for both therapists and
families in terms of treatment outcomes and child safety.
The reasons why therapists do not always perform pre-and post-
treatment outcome evaluations, remain unknown. But previous research
suggests that even though SSP is considered the gold standard for
measuring attachment (Bernard et al., 2012; Tryphonopoulos et al.,
2014), there are several challenges with using SSP evaluations in clinical
practice (Nielsen et al., 2020; Tryphonopoulos et al., 2014), such as the
resources needed for administrating the procedure and the complexity
of the analysis, which was in line with our ndings. Nevertheless,
whether these challenges are involved in therapists’ decisions not to use
SSP for all pre-and post-evaluations, is not clear from our ndings.
COS-VF uses the Secure Base-Safe Haven coding system (Marvin &
Whelan, 2007), which is not the validated version of the SSP. This is a
limitation which can offset the advantages gained by using SSP. Great
care must be taken in using measures of attachment with an intended
design to avoid, for example, adding coding-systems into a pre-existing
tool without prior validation. Thus, the way the SSP is used in COS-VF
might not give the much-needed reliable information concerning the
attachment patterns of the dyads (Tryphonopoulos et al., 2014). More-
over, because the Secure Base-Safe Haven coding manual is unpub-
lished, this impedes objective evaluations of the validity and reliability
of this coding system. As some therapists reported using COS-VF for
older children, it is important to emphasize that this also relates to the
use of measures of attachment such as SSP for younger or older ages than
it has been design for (Tryphonopoulos et al., 2014).
The questionnaire data showed that most therapists introduced the
concept of ’shark music’ to parents, and many also stated in the in-
terviews that this concept was one of the strengths of COS-VF because it
facilitated parents’ ability to reect on their unconscious defense
mechanisms and enabled a strong alliance between the parent and the
therapist. The use of components such as shark music, aligns with
ndings that indicate that labelling strong emotional responses may
help reducing the response of amygdala and other parts of the limbic
system responsible for alerting us to danger (Brooks et al., 2017; Lie-
berman et al., 2007; Lieberman et al., 2011). It has been argued that this
intensies activity in the prefrontal cortex, which can help regulate
negative affect. In other words, the “feeling” and the “thinking” parts of
the brain start communicating, which may slow down processing of
procedural memories enough for parents to reect on how it affects
them and adjust their intended/usual reaction (Powell et al., 2014).
Thus, the use of shark music in therapy seems to be a strength with, not
only COS-VF, but COS models in general and potentially other
attachment-based interventions that use similar concepts such as “voices
from the past” in Attachment and Biobehavioral Catch-Up (ABC; Ber-
nard et al., 2012).
There are several other similarities between COS-VF and ABC. For
instance, “the top of the circle” in COS, is referred to as “follow the lead”
in ABC and “the bottom of the circle”, which ABC refers to as “nurtur-
ance”. However, ABC has some core elements that are not included in
COS-VF, such as direct parental skills training with the child present in
the room. Several COS-VF therapists did, however, think that including
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
8
the children in sessions, should become a part of intervention.
Furthermore, as research on ABC suggests the intervention is effective
for at-risk children (Bernard et al., 2012; Caron et al., 2016; Dozier et al.,
2008), it might be favorable to include such core intervention compo-
nents in future developments of COS-VF and other parenting programs.
Therapists argued that experience facilitated exible use of the
intervention and the ability to be “present” in therapy sessions with the
families. Previous research suggests that exibility in using a method is
more prevalent among experienced therapists (Hubbert et al., 2001;
Tschuschke et al., 2015), and it has been proposed that exibility within
the boundaries of the treatment protocol can be related to better out-
comes (Hubbert et al., 2001). Findings from this study seem to imply
that therapists with more experience with the manual, may be more
sensitive to the parents, because uncertainties about what to say or do,
does not interfere with their ability to be “present” during sessions.
There was feedback concerning the length of the intervention in this
study. Therapists argued that COS-VF was too long to be eligible for
severe cases where quick changes were needed to ensure the safety of
the children. Though therapists have argued that there are no quick xes
to the problems COS-VF is supposed to remedy (Nielsen et al., 2020),
previous research on attachment-based interventions suggest that “less
is more”. In their meta-analysis, Bakermans-Kranenburg and colleagues
(2003) found that interventions with a small to moderate number of
sessions were more effective than interventions with more sessions.
Thus, shortening COS-VF could facilitate the implementation process as
shorter interventions make implementation more feasible (Bernard
et al., 2012), while achieving the same or even better results.
Therapists reported challenges related to the target population. The
treatment manual itself states that parents with moderate to severe
cognitive impairment will most likely not fully benet from the inter-
vention, which can be considered a rather vague description. COS-I has
items related to reective functioning; however, this is not a validated
measure. Neither does the manual describe any specic criteria for in-
clusion or exclusion of parents based on their reective functioning, as
measured by COS-I. Furthermore, the results indicated that therapists do
not learn how to adequately assess reective functioning using this
interview. Thus, this may need to receive more focus in future research,
training, and development of the intervention.
The results indicate that it is important to have realistic expectations
of the intervention, and that COS-VF by itself cannot work wonders.
Some of the challenge’s therapists experienced suggested that COS-VF
may not be suitable for everyone. Previous research shows that it is
important to examine for whom interventions work and for whom they
do not, so dissemination efforts can be targeted towards those families
(Bernard et al., 2012; Bick & Dozier, 2013). Additionally, since a pre-
vious study on COS-P suggested the intervention made some families
worse (Cassidy et al., 2017), it is important to critically address the issue
of “do no harm”, to make sure interventions such as COS-VF, do not pose
any risk or threat to anyone who receives treatment. This should always
be the main priority of any intervention and evidence-based practice.
4.1. The diffusion of COS-VF
When viewing our nding in relation to Roger’s diffusion of inno-
vation theory (Rogers, 2002) it seems that; rst, therapists see the
relative advantage of the intervention. COS-VF offers a tangible approach
to attachment theory that is comprehensive and accessible to families,
and effective in helping the families with their challenges. However,
because therapists reported that some leaders appear to be more hesitant
in making such conclusions without empirical evidence, this can be
considered a weakness and barrier to the implementation process.
Second, our ndings did not indicate any inconsistencies with the
needs and values of services and COS-VF, although there were concerns
about the safety of the children. Nevertheless, as mentioned by some
therapists, the manual can be used exibly. Thus, many had taken
measures to meet the children more often, even if it was not specied in
the manual.
Third, there were reported issues related to the complexity of COS-VF,
such as the signicant time and resources that are needed to use the
intervention as intended. This can be seen as a weakness, though there
might not be any easy solutions. This may be the best explanation as to
why attachment-based treatments are not readily embraced within child
welfare organizations (Blome et al., 2010).
Fourth, according to diffusion of innovation theory, the trialability of
COS-VF and other interventions will increase the likelihood of its
implementation. This suggests that an organization will embrace the
proposed model more readily if the intervention is implemented exibly
and on a limited basis (Rogers, 2002). Yet, even though therapists
described the manual itself as exible, the complexity of the interven-
tion requires the organization to adapt in ways that are less exible and
limited (i.e., COS-VF training, acquiring technical equipment, SSP
rooms, etc.). This can indeed be a great challenge that may affect the
implementation of COS-VF and other interventions.
4.2. Practical Implications
Findings of this study highlight several opportunities for building on
the perceived strengths and overcoming the perceived weaknesses of
COS-VF. The rst and foremost step is to perform a solid examination of
the effectiveness of COS-VF, as there are no such studies yet. Further-
more, to ensure proper assessment of attachment patterns, COS-VF
therapists should either use the SSP as it was designed, or efforts
should be made to validate the Secure Base-Safe Haven coding system
(Marvin & Whelan, 2007). To do this, the unpublished manual should be
made available for researchers, to allow independent evaluations of the
coding system. Additionally, there should be an evaluation to see if the
intervention is effective on older children as some therapists have
experienced. To this end, a measure of attachment for older children
would need to be implemented in the protocol to provide reliable as-
sessments of this age group.
The ndings suggested that therapists found it hard to assess
reective functioning. Training could thus focus more on how to assess
reective functioning using for example COS-I. However, if COS-I is not
considered to include relevant and/or enough items related to reective
function to make a proper assessment, another, preferably validated,
tool should be included in the treatment manual. Furthermore, as some
stated that the intervention was not suitable for parents with low
reective functioning, while others disagreed, there should be an eval-
uation determining the suitability of these parents. If results indicate
that these parents are not able to achieve the same progression and
outcomes as other parents, future research should evaluate whether
video-based circle stories of parent–child interactions could facilitate
the use of COS-VF these families.
To enhance safety of the children, program providers should
consider ways of including the children in therapy by, for example,
implementing home-visits as part of the intervention. Therapists can use
this as an opportunity to make parent–child interaction videos in a more
natural setting, which is a way ensuring unbiased circle stories are being
used in therapy sessions. This can also decrease the threat of parental
pseudo-mentalization. Furthermore, program providers should examine
the possibility of reducing/shortening some aspects of the intervention
to increase the feasibility of proper implementation in different mental
health and social services.
4.3. Limitations
A limitation in this study was that not everyone who were inter-
viewed responded to the questionnaire and not everyone that lled in
the questionnaire were interviewed. However, we were able to get rich,
in-depth information from the informants who participated, so there
was more than enough data to perform a proper evaluation of the COS-
VF intervention for the purposes of this study. Fifty-four percent of COS-
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
9
VF therapists in Norway at the time of the study decided not to partic-
ipate. It is difcult to speculate why therapists declined to participate,
but it seems likely that time restrictions could play a part in this, as
previous research has indicated that a lack of time is as a major chal-
lenge for many therapists (Nielsen et al., 2020). However, even though
the number of participants included in this study is low as compared to
many purely quantitative studies, it was higher than the average number
found in most qualitative studies (Levitt et al., 2017). The inclusion of
both qualitative and quantitative data also enrichened the analysis even
further. There are numerous advantages of self-report measures
(Martinez et al., 2014); however, there are also disadvantages than can
affect the results, such as social desirability bias, leniency bias, and an
individual’s mood (Podsakoff et al., 2003; Squires et al., 2011). It is
difcult to assess whether results are affected by such factors, which
may be considered a limitation of our study. Regarding the qualitative
data, there is also potential for bias, given that the rst author conducted
and transcribed all interviews, edited the transcripts, and was the lead
coder.
5. Conclusion
This study found that therapists were, generally, positive about COS-
VF and its’ outcomes. They also argued that the COS-developers have
made complex attachment theory tangible for both therapists and par-
ents. Furthermore, they stated that the assessment tools used in COS-VF
provided in-depth information that quickly and efciently got to the
root of the families’ struggles. Teaching parents the COS-language
facilitated communication between therapists and parents. However,
therapist also discussed the threat of parents developing a pseudo-
language, without integrating the new skills in their daily behavior.
Therapists considered it a weakness that COS-VF did not include more
sessions with children, which, in turn, threatened their ability to ensure
the welfare of the child. Thus, they recommended making home-visits
and additional videos of parent–child interaction as a part of the inter-
vention. Therapists also wanted to attain assessment tools that could
facilitate evaluations of parents reective functioning. As COS-VF is not
yet evidence-based, there is thus a need for evaluations of the effec-
tiveness of COS-VF in terms of a randomized trial, before continuing
training and future scale-up. This is critical to assure families get the
help they need and to ensure the safety of children.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
The data that has been used is condential.
References
Aber, J. L., Slade, A., Berger, B., Bresgi, I., & Kaplan, M. (1985). The parent development
interview: Manual and administration procedures. New York: Barnard College,
Department of Psychology.
Abidin, R. R. (1995). Parenting Stress Index (PSI). manual ((3rd ed.).). Charlottesville: VA:
Pediatric Psychology Press.
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment.
NJ Eribaum: Hills-dale.
Andrews, E., & Coyne, J. (2018). Travelling the circle together, solo : An individual
protocol for the Circle of Security intensive intervention, 1–13. https://doi.org/
10.1002/jclp.22644.
Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Juffer, F. (2003). Less is more:
Meta-analyses of sensitivity and attachment interventions in early childhood.
Psychological Bulletin, 129(2), 195–215. https://doi.org/10.1037/0033-
2909.129.2.195.
Berlin, L. J., Zeanah, C. H., & Liberman, A. F. (2016). Prevention and Intervention
Programs to Support Early Attachment Security: A Move to the Level of the
Community. In J. Cassidy, & P. R. Shaver (Eds.), Handbook of Attachment ((3rd ed.,,
pp. 739–758). Guilford Press.
Bernard, K., Dozier, M., Bick, J., Lewis-Morrarty, E., Lindhiem, O., & Carlson, E. (2012).
Enhancing Attachment Organization Among Maltreated Children: Results of a
Randomized Clinical Trial. Child Development, 83(2), 623–636. https://doi.org/
10.1111/j.1467-8624.2011.01712.x
Bernstein, W. M., Stephan, W. G., & Davis, M. H. (1979). Explaining attributions for
achievement: A path analytic approach. Journal of Personality and Social Psychology,
37, 1810–1821. https://doi.org/10.1037/0022-3514.37.10.1810.
Bick, J., & Dozier, M. (2013). The effectiveness of an attachment-based intervention in
promoting foster mothers’ sensitivity toward foster infants. Infant Mental Health
Journal, 34(2), 95–103. https://doi.org/10.1002/imhj.21373
Blome, W. W., Bennett, S., & Page, T. F. (2010). Organizational Challenges to
Implementing Attachment-Based Practices in Public Child Welfare Agencies: An
Example Using the Circle of Security® Model. Journal of Public Child Welfare, 4(4),
427–449. https://doi.org/10.1080/15548732.2010.526904
Brooks, J. A., Shablack, H., Gendron, M., Satpute, A. B., Parrish, M. H., & Lindquist, K. A.
(2017). The role of language in the experience and perception of emotion: A
neuroimaging meta-analysis. Social Cognitive and Affective Neuroscience, 12(2),
169–183. https://doi.org/10.1093/scan/nsw121
Caron, E. B., Weston-Lee, P., Haggerty, D., & Dozier, M. (2016). Community
implementation outcomes of Attachment and Biobehavioral Catch-up. Child Abuse
and Neglect, 53, 128–137. https://doi.org/10.1016/j.chiabu.2015.11.010
Cassidy, J., Brett, B. E., Gross, J. T., Stern, J. A., Martin, D. R., Mohr, J. J., &
Woodhouse, S. S. (2017). Circle of Security-Parenting: A randomized controlled trial
in Head Start. Development and Psychopathology, 29(2), 651–673. https://doi.org/
10.1017/S0954579417000244
Cassidy, J., & Marvin, R. S. (1992). Attachment organization in preschool children:
Procedures and coding manual.
Creswell, J. W., Plano-Clark, V. L., & Garrett, A. L. (2008). Methodological Issues in
Conducting Mixed Methods Research Designs. In M. M. Bergman (Ed.), Advances in
mixed methods research: Theories and applications (1st ed., pp. 67–83). London: SAGE
Publications Ltd.
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of
psychopathology. Journal of Consulting Psyle:///C:/Users/Niels/Downloads/
C19b01b04ca7dbf122d59684bf05353cc77b(1). Pdfchology, 24, 349–354.
Dehghani, A., Malekpour, M., Abedi, A., & Amiri, S. (2014). The efcacy of circle of
security on attachment and well-being in preschool children. Management Science
Letters, 4, 607–612. https://doi.org/10.5267/j.msl.2013.12.031
Derogatis, L. R., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., & Covi, L. (1974). The
Hopkins Symptom Checklist (HSCL): A self-report symptom inventory. Behavioral
Science, 19(1), 1–15. https://doi.org/10.1002/bs.3830190102
Dozier, M., Peloso, E., Lewis, E., Laurenceau, J. P., & Levine, S. (2008). Effects of an
attachment-based intervention on the cortisol production of infants and toddlers in
foster care. Development and Psychopathology, 20(3), 845–859. https://doi.org/
10.1017/S0954579408000400
Fraiberg, D., Adelson, E., & Shapiro, V. (1975). Ghost in the Nursery: A Psychoanalytic
Approach to the Problems of Impaired Infant-Mother Relationships. The Journal of
the American Academy of Child and Adolescent Psychiatry, 14(3), 387–421.
George, C., Kaplan, N., & Main, M. (1996). Adult Attachment Interview (Unpublished
manuscript).
Helms, M. M., & Nixon, J. (2010). Exploring SWOT analysis – where are we now?: A
review of academic research from the last decade. In. Journal of Strategy and
Management, 3(3). https://doi.org/10.1108/1755425101106483
Hoffman, K. T., Marvin, R. S., Cooper, G., & Powell, B. (2006). Changing Toddlers’ and
Preschoolers’ Attachment Classications: The Circle of Security Intervention. Journal
of Consulting and Clinical Psychology, 74(6), 1017–1026. https://doi.org/10.1037/
0022-006X.74.6.1017
Hubbert, J. D., Bufka, L. F., Barlow, D. H., Gorman, J. M., Shear, M. K., & Woods, S. W.
(2001). Therapists, Therapist Variables, and Cognitive-Behavioral Therapy
Outcomes in a Multicenter Trial for Panic Disorders. Journal of Counsulting and
Clinical Psychology, 69(5), 747–755. https://doi.org/10.1037/0022-006X.69.5.747.
Huber, A., McMahon, C. A., & Sweller, N. (2015). Efcacy of the 20-week circle of se-
curity intervention: Changes in caregiver reective functioning, representations, and
child attachment in an Australian clinical sample. Infant Mental Health Journal, 36
(6), 556–574. https://doi.org/10.1002/imhj.21540
Letourneau, N., Tryphonopoulos, P., Giesbrecht, G., Dennis, C.-L., Bhogal, S., &
Watson, B. (2015). Narrative and Meta-Analytic Review of Interventions Aiming to
Improve Maternal-Child Attachment Security. Infant Mental Health Journal, 36(4),
366–387. https://doi.org/10.1002/imhj.21525
Levitt, H. M., Pommerville, A., Surace, F. I., & Grabowski, L. M. (2017). Metamethod
study of qualitative psychotherapy research on clients’ experiences: Review and
recommendations. Retrieved from Journal of Counseling Psychology, 64(6), 626–644
https://psycnet.apa.org/fulltext/2017-51520-004.pdf.
Lieberman, M. D., Eisenberger, N. I., Crockett, M. J., Tom, S. M., Pfeifer, J. H., &
Way, B. M. (2007). Putting Feelings Into Words. Psychological Science, 18(5),
421–428. https://doi.org/10.1111/j.1467-9280.2007.01916.x
Lieberman, M. D., Inagaki, T. K., Tibibnia, G., & Crockett, M. J. (2011). Subjective
responses to emotional stimuli during labeling, reappraisal, and distraction. Emotion,
468–480. https://doi.org/10.1037/a0023503
Lyon, A. R., & Koerner, K. (2016). User-Centered Design for Psychosocial Intervention
Development and Implementation. Clinical Psychology: Science and Practice, 23(2),
180–200. https://doi.org/10.1111/cpsp.12154
Martinez, R. G., Lewis, C. C., & Weiner, B. J. (2014). Instrumentation issues in
implementation science. Implementation Science, 9(1), 1–9. https://doi.org/10.1186/
s13012-014-0118-8
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
10
Marvin, R. S., Cooper, G., Hoffman, K. T., & Powell, B. (2002). The Circle of Security
project : Attachment-based intervention with caregiver – pre-school child dyads.
Attachment & Human Development, 4(1), 107–124. https://doi.org/10.1080/
1461673021013163
Marvin, R. S., & Whelan, W. F. (2006). Trygghetssirkelen: Manualen for
Familiemodellen. Oslo: Nasjonalt kompetansesenter for barn og unges psykiske
helse. Regionsenter for barn og unges psykiske helse, Helseregion Øst og Sør.
Marvin, R. S., & Whelan, W. F. (2007). Secure Base Safe Haven Coding System: Unpublished
manual. USA: Virginia.
Maupin, A. N., Samuel, E. E., Nappi, S. M., Heath, J. M., & Smith, M. V. (2017).
Disseminating a Parenting Intervention in the Community: Experiences from a Multi-
Site Evaluation. Journal of Child and Family Studies, 26(11), 3079–3092. https://doi.
org/10.1007/s10826-017-0804-7
Mountain, G., Cahill, J., & Thorpe, H. (2017). Sensitivity and attachment interventions in
early childhood: A systematic review and meta-analysis. Infant Behavior and
Development, 46, 14–32. https://doi.org/10.1016/j.infbeh.2016.10.006
Nielsen, B., Oddli, H. W., Slinning, K., & Drozd, F. (2020). Implementation of
Attachment-Based Interventions in Mental Health and Social Welfare Services:
Therapist’s Experiences from the Circle of Security-Virginia Family Intervention.
Children and Youth Services Review, 108, 1–14. https://doi.org/10.1016/j.childyouth.
2019.104550.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common Method
Biases in Behavioral Research: A Critical Review of the Literature and Recommended
Remedies. Journal of Applied Psychology, 88(5), 879–903. https://doi.org/10.1037/
0021-9010.88.5.879
Powell, B., Cooper, G., Hoffman, K., & Marvin, B. (2014). The Circle of Security
Intervention: Enhancing Attachment in Early Parent-child Relationships. New York:
Guilford Press.
Ramsauer, B., Mühlhan, C., Lotzin, A., Achtergarde, S., Mueller, J., Krink, S., & Romer, G.
(2019). Randomized controlled trial of the Circle of Security-Intensive intervention
for mothers with postpartum depression: Maternal unresolved attachment moderates
changes in sensitivity. Attachment and Human Development, 00(00), 1–22. https://
doi.org/10.1080/14616734.2019.1689406
Rogers, E. M. (2002). Diffusion of preventive innovations. Retrieved from Addictive
Behaviors, 27(6), 989–993 http://www.sciencedirect.com.miman.bib.bth.se/scien
ce/article/pii/S0306460302003003.
Squires, J. E., Hutchinson, A. M., Bostr¨
om, A. M., O’Rourke, H. M., Cobban, S. J., &
Estabrooks, C. A. (2011). To what extent do nurses use research in clinical practice?
A systematic review. Implementation Science, 6(1), 1–17. https://doi.org/10.1186/
1748-5908-6-21
Tracey, T. J. G., Wampold, B. E., Lichtenberg, J. W., & Goodyear, R. K. (2014). Expertise
in psychotherapy: An elusive goal? American Psychologist, 69(3), 218–229. https://
doi.org/10.1037/a0035099
Tryphonopoulos, P. D., Letourneau, N., & Ditommaso, E. (2014). Attachment and
caregiver-infant interaction: A review of observational-assessment tools. Infant
Mental Health Journal, 35(6), 642–656. https://doi.org/10.1002/imhj.21461
Tschuschke, V., Crameri, A., Koehler, M., Berglar, J., Muth, K., Staczan, P., & Koemeda-
Lutz, M. (2015). The role of therapists’ treatment adherence, professional
experience, therapeutic alliance, and clients’ severity of psychological problems:
Prediction of treatment outcome in eight different psychotherapy approaches.
Preliminary results of a naturalistic. Psychotherapy Research, 25(4), 420–434.
https://doi.org/10.1080/10503307.2014.896055
Walsh, S., McAlister, B., O’donnell, P., & Lambert, M. J. (2012). An investigation of self-
assessment bias in mental health Providers. Psychological Reports, 110(2), 639–644.
https://doi.org/10.2466/02.07.17.PR0.110.2.639-644.
Wright, B., Barry, M., Hughes, E., Tr´
epel, D., Ali, S., Allgar, V., & Gilbody, S. (2015).
Clinical effectiveness and cost-effectiveness of parenting interventions for children
with severe attachment problems: A systematic review and meta-analysis. Health
Technology Assessment (Winchester, England), 19(52), v1–347. https://doi.org/
10.3310/hta19520
Zeanah, C. H., Benoit, D., Hirshberg, L., Barton, M. L., & Regan, C. (1994). Mothers’
representations of their infants are concordant with infant attachment
classications. Developmental Issues in Psychiatry and Psychology, 1, 9–18.
B. Nielsen et al.
Children and Youth Services Review 169 (2025) 108048
11
