Introduction
Coercive control
Coercive control (CC) is increasingly recognised as the defining feature of intimate partner violence (IPV), expanding the conceptualisation of victim experiences beyond the traditional focus on physical and sexual abuse (Johnson, Reference Johnson1995; Schechter, Reference Schechter1982; Stark, Reference Stark2007). Understanding IPV through this lens shifts attention from isolated incidents of violence to the sustained, strategic patterns of control that define abusive relationships (Anderson, Reference Anderson2008; Lischick, Reference Lischick1999; Lischick, Reference Lischick, Stark and Buzawa2009).
Stark (Reference Stark2007) understands coercion as the use of threats to achieve a desired response from the survivor and control as structural forms of deprivation, exploitation, and commands that indirectly gain the survivor’s obedience. In the presence of both these constructs together, he argues the survivor experiences a ‘condition of unfreedom’ or entrapment. Thus, CC is conceptualised as the use of primarily non-violent strategies (isolation, monitoring, etc.) to maintain dominance over one’s partner in an intimate relationship.
CC has been shown to be both more frequent and severe than other IPV forms, and often persists post-separation (Johnson, Reference Johnson2008; Ornstein and Rickne, Reference Ornstein and Rickne2013). It can also precede, motivate, or increase the likelihood of other forms of violence in relationships (Graham-Kevan and Archer, Reference Graham-Kevan and Archer2008; Hardesty et al., Reference Hardesty, Crossman, Haselschwerdt, Raffaelli, Ogolsky and Johnson2015), and the dynamics and impact of CC without violence appear similar to those of CC with violence (Crossman et al., Reference Crossman, Hardesty and Raffaelli2016). IPV, including CC, adversely affects economic, physical and mental health outcomes (Bosch et al., Reference Bosch, Weaver, Arnold and Clark2017; Lohmann et al., Reference Lohmann, Cowlishaw, Ney, O’Donnell and Felmingham2023).
Although CC is also experienced by men and in same-sex couples (Raghavan et al., Reference Raghavan, Beck, Menke and Loveland2019; Stark and Hester, Reference Stark and Hester2019), it is a gendered construct (meaning women more than men are likely to experience the phenomenon). Ninety-five per cent of CC survivors in UK police data were women, and 74% of perpetrators were men (Barlow et al., Reference Barlow, Walklate, Johnson, Humphreys and Kirby2018). Data from the Crime Survey for England and Wales also highlighted that women are far more likely than men to be victims of abuse that involves ongoing degradation and frightening threats – the two key elements of CC (Myhill, Reference Myhill2015).
Despite there being no nationally agreed measure of CC in the UK currently, rates of recorded offences of the crime offer helpful insights into its likely prevalence rate. In the year ending March 2022, domestic abuse-related crimes represented 17.1% of all offences recorded by the police (Office for National Statistics, 2022) and it is known that CC is the most common of all forms of IPV (Kanougiya et al., Reference Kanougiya, Sivakami and Rai2021). Rates of recorded offences of CC are increasing significantly annually (43,774 in the year ending March 2023, compared with 9053 in 2018; Office for National Statistics, 2023). It is likely that the increasing awareness of the law is contributing to more survivors filing reports and the true prevalence and pervasiveness of CC being known. These significant numbers of recorded crimes highlight the need for CC to become a research and policy priority. In the USA, approximately one in three women (39.7%) have experienced CC by an intimate partner in their lifetime (Smith et al., Reference Smith, Zhang, Basile, Merrick, Wang, Kresnow and Chen2018). In a longitudinal study conducted by Solace Women’s Aid with women and children rebuilding their lives after experiencing domestic violence, 95 out of the 100 IPV survivor participants reported experiencing CC (Kelly et al., Reference Kelly, Sharp-Jeffs and Klein2014).
However, due to the historical understanding of IPV as discrete events of violence, most IPV assessments in healthcare and other settings focus primarily on acts of violence without including elements of coercion (Dutton and Goodman, Reference Dutton and Goodman2005). Stark and Hester (Reference Stark and Hester2019) discussed that when professionals do not understand CC, many chronic offenders remain undetected because professionals are trained to assess individual acts of violence, which may seem minor by themselves until understood through the context of ongoing CC. Crossman and Hardesty (Reference Crossman and Hardesty2018) also found that in cases of CC where violence is not present, women face barriers to identifying and accessing support as professionals and policies may minimise their needs.
Psychological model and intervention for CC
CC has been explained as a construct as part of several IPV models and in some CC-specific models (Dutton and Goodman, Reference Dutton and Goodman2005; Okun, Reference Okun1986; Stark, Reference Stark2007). Varied psychological theories, such as cognitive behavioural, trauma-informed, and psychopathology-stress models, exist to explain the aetiology of IPV perpetration (Eckhardt and Massa, Reference Eckhardt, Massa, Geffner, White, Hamberger, Rosenbaum, Vaughan-Eden and Vieth2022), but none until Ager (Reference Ager2020) focused on survivor behaviours in IPV or discussed CC. Ager interviewed 12 women survivors of IPV in the USA and developed a cognitive behavioural model of IPV. His theory highlighted the antecedents of violent situations in IPV couples (including life circumstances, perpetrator control over survivor, triggers, etc.), relevant factors during the IPV event (feelings, behaviours, survivor cognitions and prevention activities, etc.) and after the event (survivor cognitions and reactions, third-party involvement, reconciliation, etc.). While this approach was novel, the results of his analysis view IPV primarily as physical violence and discrete events, which is not inclusive of the sustained and patterned nature of CC (although the model does discuss control as an antecedent factor to violence). Thus, while existing theories on IPV and CC discuss the causes and factors that perpetuate them, especially from the perpetrator’s perspective, and the resulting impacts on the survivor, there is no specific psychological model and therefore intervention for CC yet.
Existing IPV interventions focus on managing specific symptoms that survivors experience, such as low mood and anxiety, formulated using psychological models. Cognitive behavioural therapy (CBT; American Psychological Association, 2017) has the largest evidence base in this regard, and has been used effectively to work with IPV survivors to manage depression and low self-esteem (de los Angeles Cruz-Almanza et al., Reference de los Angeles Cruz-Almanza, Gaona-Márquez and Sánchez-Sosa2006). In a randomised clinical trial with women survivors of IPV, Johnson and colleagues (Reference Johnson, Zlotnick and Perez2011) found reductions in post-traumatic stress disorder (PTSD) symptoms and the likelihood of re-abuse with a CBT intervention versus standard shelter services. These gains continued at a 6-month follow-up. A meta-analysis of short-term therapies for IPV also found that CBT interventions tailored to IPV survivors showed the greatest improvements (Arroyo et al., Reference Arroyo, Lundahl, Butters, Vanderloo and Wood2017).
Considering CBT’s evidence base in work with IPV survivors, it is likely that a CBT-based understanding of CC may be helpful in formulating survivor experiences and tailoring interventions based on their specific experiences. Thus, the aim of the current study was to develop a CBT-informed psychological model of CC focusing on women’s experiences.
Method
Design
The study adopted a cross-sectional, qualitative, semi-structured interview design, with constructivist grounded theory (GT; Charmaz, Reference Charmaz2000) methodology.
Grounded theory
An exploratory qualitative design was employed for this study to analyse interview data, with GT methodology (Corbin and Strauss, Reference Corbin and Strauss2015; Glaser and Strauss, Reference Glaser and Strauss1967) selected for its primary goal of developing a theoretical framework that explains phenomena, rather than merely describing them (Willig, Reference Willig2001). Grounded theory is particularly well-suited to studies seeking to generate new theoretical insights, and it is an ideal approach for exploring complex, under-researched topics like CC. GT is designed to identify and link processes at multiple levels: individual, interpersonal, and broader contextual and systemic factors. This multi-level focus aligns with the aim of this study to develop a comprehensive theory that captures the psychological experiences of women survivors of coercive control, reflecting the ‘average’ person within the data while also accounting for variation and nuance.
Furthermore, GT allows for the extension of existing theoretical frameworks. In this study, constructivist GT (Charmaz, Reference Charmaz2006; Charmaz, Reference Charmaz2014) was specifically chosen because it enables the integration of pre-existing theories, such as CBT, into the theory-building process. Constructivist GT recognises the interaction between researchers and participants in the generation of knowledge and allows for flexibility in incorporating established theories to inform the data collection and analysis process. By adopting CBT as a guiding framework for interview questions and data categorisation, this study not only contributes to theory-building in the domain of CC but also seeks to bridge the gap between this emerging theory and existing knowledge and therapeutic practices.
Reflexivity
The first author is a female trainee clinical psychologist, which influenced her research interest in CC, as she was aware of its gendered nature prior to conducting this research. The second author is a clinical psychologist with extensive experience of delivering CBT interventions and was the research supervisor for this study. While constructivist GT acknowledges that the researcher can have prior knowledge about the phenomenon under study (Charmaz, Reference Charmaz2014), it is important that existing information does not prevent researchers from seeing things that do not fit with their pre-conceived notions (Corbin and Strauss, Reference Corbin and Strauss2008). Thus, to minimise that risk, the first author wrote memos (Chun Tie et al., Reference Chun Tie, Birks and Francis2019) throughout the study and reflected on her assumptions during data collection. Supervision and peer reflective spaces were also helpful in thinking through difficult codes or material coming up in interviews that did not fit with her prior knowledge.
Participants
Inclusion and exclusion criteria
The inclusion criteria for the study were: adult women with self-identified experience of CC with an intimate partner and fluency in English. The exclusion criteria were: current experience of a condition that could affect their sense of reality (e.g. psychosis), dependence on illicit drugs or alcohol, currently self-harming or experiencing suicidal ideation (due to the study being a doctorate research project and no having sufficient resources to manage risk), living in a situation of risk (e.g. living with their perpetrator in an environment of ongoing abuse) or being unable to provide informed consent.
Sample
Fourteen women (n=14, mean age 45.9 years; Table 1) were recruited through online social media platform X (previously Twitter). All participants self-identified as having experienced CC in heterosexual relationships with men. Sampling was purposive and followed GT methodology of theoretical sampling after initial interviews are analysed; however, theoretical saturation could not be fully ensured due to time pressures as this was a doctoral project with strict deadlines. All women who self-referred to the study were eligible for participation. In terms of ethnicity, the majority of the participants were white British (n=6), followed by white-other participants (individuals who identify as white but are not of English, Welsh, Scottish, Irish, or Irish Traveller heritage; n=2). Amongst the remaining participants, there was one each of white Scottish, British Indian, British Pakistani, Asian-other (individuals who self-identify as having an Asian heritage but do not fall under the categories of British Indian, British Pakistani, British Bangladeshi, or British Chinese), Arab, and black Caribbean ethnicity.
Materials and measures
Checklist of controlling behaviours
All participants completed the Checklist of Controlling Behaviours (CCB) questionnaire (Lehmann et al., Reference Lehmann, Simmons and Pillai2012; see Appendix A in the Supplementary material) as a screening tool prior to the interview. Based on the questionnaire results (Table 1), 57% of the participants had experienced frequent or very frequent CC, while 43% had experienced occasional CC. None experienced little or rare CC (score range of 0–164).
Semi-structured interview
The first version of the semi-structured interview schedule (see Appendix B in the Supplementary material) was developed with the second author and two experts by experience (EBE; individuals with experience of CC who provided feedback on the design of the study). Interview schedules were analysed following each interview and adapted, if necessary, based on the GT approach (Charmaz, Reference Charmaz2014) if new material relevant to theory-building came up. Adapting interview schedules was also important in utilising a form of theoretical sampling, whereby adaptive interview schedules steered questions in the direction of emergent theorising, allowing for theoretical sufficiency (Conlon et al., Reference Conlon, Timonen, Elliott-O’Dare, O’Keeffe and Foley2020, p. 949; Foley et al., Reference Foley, Timonen, Conlon and O’Dare2021).
Procedure
Participants registered interest in the study by emailing the researcher, and were sent the participant information sheet and booked in for an introductory telephone call (to clarify any queries, ascertain they met the inclusion criteria, complete the CCB questionnaire, and collect demographic data). If participants revealed information indicating any form of current risk, they were excluded from the study. They were informed of the tentative month their interview would take place. Interviews were conducted online on Microsoft Teams, and consisted of a 5-minute introduction to reiterate the aim of the study, confidentiality, and encourage participants to ask for breaks if needed. Informed consent was sought, followed by interviews lasting 60–75 minutes, and a 5-minute debrief. Participants were sent a debrief form after the interview with information about support and crisis services they could access. After data analysis was completed, the initial model was sent to the participants via a Qualtrics survey for their feedback. Seven participants responded to the survey; all participants thought the model was reflective of their experiences and comprehensive. Two participants provided feedback on elaborating the meanings of certain cognitions and coping mechanisms, which were incorporated into the final model.
Data analysis
Each interview transcript was analysed before the subsequent interview took place, to allow for the data analysis to inform the next data collection point. Initial codes (Mills et al., Reference Mills, Birks, Hoare, Mills and Berks2014) were first developed, whereby each sentence (or a small group of sentences) was labelled to make meaning of it in relation to the research question. The next stage, focused coding (Mills et al., Reference Mills, Birks, Hoare, Mills and Berks2014), involved developing categories from the initial codes. Throughout the analysis, a constant comparative method was employed by systematically comparing new data with existing codes and categories, enabling the refinement of emerging themes and the development of a coherent, data-driven theoretical framework. Relationships between categories were considered and recruitment stopped after the analysis of the final two interviews did not add new codes (Birks and Mills, Reference Birks and Mills2015). The final stage was theoretical coding (Mills et al., Reference Mills, Birks, Hoare, Mills and Berks2014), where the final grounded theory was integrated by developing a set of interrelated concepts that linked the focused codes (Chun Tie et al., Reference Chun Tie, Birks and Francis2019).
Trustworthiness
Lincoln and Guba (Reference Lincoln and Guba1985) presented four criteria to develop trustworthiness in qualitative research: credibility, dependability, confirmability, and transferability. The design of the study was carefully considered to prioritise achieving the four criteria as much as possible. GT methodology, which is well established in qualitative investigation, was used to increase credibility in the study design. Theoretical sampling is integral to GT methodology and adapting interview schedules helped to achieve a form of that, contributing to theoretical sufficiency (Conlon et al., Reference Conlon, Timonen, Elliott-O’Dare, O’Keeffe and Foley2020, p. 949; Foley et al., Reference Foley, Timonen, Conlon and O’Dare2021). It was hoped that the support of a second reviewer in independently going through a sample of the interview transcripts, resulting codes, and the emerging findings would contribute to the dependability of the study. Similarly, member-checking, memo-ing, and regular supervision supported with acknowledging and reducing researcher biases. The emergent model from this study can be replicated in future studies with larger samples and diverse groups of women, which helps with its transferability.
Results
Six theoretical codes emerged from the data, consisting of 32 focused codes (Fig. 1). While there may have been other maintaining factors resulting from perpetrator behaviours, the current findings only reflect survivor perspectives. Figure 1 provides an integrative depiction of these six theoretical codes within a cognitive-behavioural framework. Reading the diagram clockwise, the model proposes that pre-existing vulnerability factors feed into survivors’ core cognitions (e.g. ‘I deserve the abuse’) which, consistent with CBT theory, then influences survivors’ emotional states (shame, fear, sadness, anger). These emotional states both arise from, and reciprocally reinforce, a set of maintaining factors – perceived consequences of leaving, restricted finances, damaged self-worth and constraining social norms. Survivors respond with two forms of behaviour: short-term unhelpful coping (dissociation, avoidance, subjugation, substance use) and protective factors that buffer distress (social connection, valued activity, physical exercise). The cumulative impact of these interacting processes manifests in deteriorating mental and physical health, mistrust and economic hardship (impact). Together, the cyclical arrows in Fig. 1 illustrate how CC is enacted and perpetuated through an interlocking system of thoughts, emotions and behaviours, rather than isolated violent episodes, thereby expanding prior content- or theme-based accounts of CC.
Figure 1. Depiction of the six theoretical codes.
Theoretical code 1: Vulnerability factors
These were the experiences prior to the CC relationship that participants felt played a role in how they related to the perpetrator: difficult childhood experiences, negative previous romantic experiences, and low self-esteem (Table 2).
Table 2. Vulnerability factors
Participants described experiences of childhood trauma or parental separation, which they associated with their later experience of CC:
‘I’d come from a very chaotic childhood. Both my parents were alcoholics, my father was violent. One minute he was there, then he would disappear off … Myself and my brother were illegitimate children of his, whom he was ashamed of … There was all of this shame around us as children and we were held responsible for shaming him … I think that’s why my childhood made me quite vulnerable to this. Always thinking that maybe somehow I got it wrong … I’ve heard people laugh and dismiss me as a child. And it made me feel that I wasn’t to be taken seriously.’ [P4]
Most participants also had negative experiences of previous romantic relationships, which ranged from infidelity to domestic abuse. Additionally, they reflected on their self-esteem being low and not valuing themselves even before the relationship.
Theoretical code 2: Cognitions
Participants in this study described seven core cognitionsthey experienced during the relationship: ‘I am not worthy’, ‘I am alone’, ‘I deserve the abuse’, ‘Is this really happening?’, ‘It might get better’, ‘He is being unfair’, and suicidal thoughts (Table 3).
Table 3. Cognitions during the relationship
Participants routinely felt worthless in their relationships and some justified the abuse through thinking that they must have done something to deserve the perpetrator’s behaviours towards them:
‘I thought I deserved it. I remember when I locked myself in the bathroom… I said, “What have I done?”. I felt responsible for causing his anger, like I should have adjusted my behaviour. Maybe I should have answered differently. And that set the precedent … if I challenged him or said no, this is the response that I would get. So it made me feel like I ought to think better for next time.’ [P2]
Some also felt that the abuse eventually would decrease with time, if they tried to be better partners to the perpetrator. All participants felt isolated during the relationship, and would often feel confused or question their sense of reality during the middle of the relationship. During this time, they also started feeling like the perpetrator was holding them to different standards than himself, which they found unfair. Most participants reflected on not wanting to live any more during this time, and experiencing suicidal thoughts.
Theoretical code 3: Affect during relationship
The participants in this study described four significant emotional states they experienced during the CC relationship: shame, fear, sadness, and anger (Table 4).
Table 4. Affects experienced during the relationship
Participants reported feeling ashamed that somebody was treating them unfairly or talking to them disrespectfully. They constantly lived in fear (of consequences, threats, sexual, and physical abuse) throughout the relationship and reflected having to ‘walk on eggshells’:
‘At the end I was living in complete control and fear to the extent that when he was sleeping, I wasn’t allowed to wake him up. I wasn’t allowed to leave the house without him knowing, so I would not sleep because of the depression and I would creep into the kitchen because the floorboards made noises and I would put a towel over the kettle so that it wouldn’t make noise. I didn’t want to wake him up, because those times between 7:00 and 11:00 were when I could avoid danger … That was the only peace I had, but I was so hypervigilant about what’s gonna happen next.’ [P6]
They experienced pervasive sadness related to hopelessness and helplessness, and some were diagnosed with depression during the relationship. Anger was experienced by majority of the participants during the middle to the end of their relationship, when they realised they were being treated unfairly.
Theoretical code 4: Maintaining factors
The participants in this study reflected on four maintaining factors that kept the abuse going and prevented them from initiating any kind of change or leaving the relationship for a long time: perceived consequences of leaving, financial situation, low self-esteem and self-worth, and social norms (Table 5).
Table 5. Maintaining factors that kept survivors in the relationship
Many participants did not wish to leave the relationship because they were fearful of the perpetrator’s threats and consequences of leaving on them (or their children and family). Participants also reflected that during the relationship, perpetrators had coerced them into leaving their jobs or reducing their working hours, so they could spend more time at home with them or the children, leading to reduced financial independence and ability to leave the perpetrator. Participants’ beliefs about their own worth and what they deserve also reduced significantly during the relationship, impacting their agency to leave:
‘During the relationship. I think it [self-esteem] got down and down. At the end … it was non-existent … I didn’t even feel like a person. I didn’t have any agency … When I thought, “Okay … I want to divorce”, it’s almost like I went and asked permission from him … Now I’m laughing because if he didn’t agree, it’s not for him to have agreed … But at the time my self-esteem and confidence was so low … I was like a little girl needing permission to do things.’ [P9]
Additionally, several participants came from communities or cultures (such as South Asian or Arab cultures) where it was shameful to be divorced or a single mother.
Theoretical code 5: Behaviours
Participants in this study broadly reported two kinds of behaviours, described below.
Unhelpful coping strategies
The coping strategies used by participants in this study were reported by them as very helpful in the short-term to survive their abusive situations, but in the long run, they were not very helpful as they did nothing to improve their situations or lives (Table 6). These included dissociation, subjugation, avoidance, and substance misuse.
Table 6. Coping strategies used by survivors during the relationship
Most participants described ‘checking out’ or ‘switching off’ from their sense of self or thoughts and feelings during difficult situations with the perpetrator. They also reported intentionally minimising their needs so they would not feel hurt when the perpetrator invalidated their feelings or expectations. Most participants coped by actively avoiding their thoughts and feelings, and trying to suppress them:
‘It was like my brain split and I lost who I was, because there is that sense of “someone just strangled me and … that’s someone who pretends to love me”. I looked into his dead eyes as he strangled me … he would do that often after that … I didn’t do anything … to be able to survive, I justified it. I stopped those thoughts. I was just like, okay, I just don’t think about it.’ [P6]
Some started using substances, in an effort to escape their distressing feelings and thoughts, and have some mental or physical separation from the perpetrator.
Protective factors
Survivors reflected that certain habits, activities, or behaviours made them feel better about themselves or brought them joy, which was a welcome presence during the abusive relationship (Table 7). These included spending time with their children, social network, and pets, engaging in physical activity, and committing to their career or interests:
‘I got a lot of, and still get huge amounts of soft affirmation … from my job. I’m really lucky I’m an <job role>… and I get a lot [mentally] from the position. It’s very rewarding and that’s what saved me actually … those were my moments of freedom, where I could really be myself.’ [P12]
Table 7. Protective factors
From a CBT perspective, these protective actions could be operating as positive behavioural experiments that momentarily disconfirm the perpetrator’s narrative of worthlessness and reduce emotional avoidance, thereby protecting against feelings of shame and hopelessness. Conceptually, they may function as moderators within the model: when engaged, they weaken the cognitive-affective loop that reinforces control, supporting survivors’ resilience and eventual disengagement from the abusive dynamic.
Theoretical code 6: Impact
Participants in the study also reported a range of experiences that described the perceived impact of the relationship on them: poor physical health, depression, anxiety, trauma, financial difficulties, and a chronic mistrust of people and institutions (Table 8):
‘Hopeless. Sometimes I felt totally hopeless. Like what is the point? … I was diagnosed with situational depression about six months after I’d left him, which lifted with the medication quite quickly, thankfully, but I was very low, tearful. Even when my friends wanted to meet up, I would find excuses not to go, because it’s not worth the battle. Why should I bother?’ [P12]
Table 8. Impacts of the relationship on the survivors
Discussion
This study articulates a preliminary cognitive behavioural model of CC grounded in women survivors’ accounts. By mapping participants’ experiences onto established CBT mechanisms, such as core beliefs, automatic thoughts, emotion regulation and behavioural responses, the model illustrates how CC may operate psychologically from survivors’ perspectives rather than merely describing what happens. Vulnerability factors (e.g. childhood adversity, prior abusive relationships) are conceptualised as formative experiences that may sensitise core beliefs of worthlessness. When triggered by the perpetrator’s controlling tactics, these core beliefs may generate maladaptive cognitions such as, ‘I deserve the abuse’, which in turn elicit distressing affect (such as shame and fear). These emotions narrow behavioural options, making avoidance, dissociation, and appeasement appear functionally adaptive in the short term and thus maintain the abusive context.
Conversely, the model highlights protective factors – activities and relationships that momentarily re-activate alternative, empowering core beliefs (such as of competence and belonging). In doing so, they buffer the impact of CC and create avenue for therapeutic change. These theory-driven links are demonstrated through dashed lines in Fig. 1, while participant-driven links are demonstrated through continuous lines. By specifying these dynamic links, the proposed model extends previous thematic analyses of CC and provides clinicians with a preliminary formulation map that could support with targeted CBT interventions.
Relevance of the findings to the literature
Risk and protective factors
A substantial body of literature addresses the risk and protective factors of IPV; however, no study to date has explored these for CC specifically. There is a significant overlap between IPV risk factors and those found for CC in this study, such as difficult childhood experiences (Cao et al., Reference Cao, Lee, Liu and Gonzalez-Guarda2021; Capaldi et al., Reference Capaldi, Knoble, Shortt and Kim2012; Gerino et al., Reference Gerino, Caldarera, Curti, Brustia and Rollè2018) and low-self esteem (Costa and Gomes, Reference Costa and Gomes2018; Papadakaki et al., Reference Papadakaki, Tzamalouka, Chatzifotiou and Chliaoutakis2008). Regarding the protective factors found in this study, social support has been the most widely researched in IPV survivors (Cao et al., Reference Cao, Lee, Liu and Gonzalez-Guarda2021; Capaldi et al., Reference Capaldi, Knoble, Shortt and Kim2012; Gerino et al., Reference Gerino, Caldarera, Curti, Brustia and Rollè2018; Howell et al., Reference Howell, Thurston, Schwartz, Jamison and Hasselle2018). A study with 36 IPV survivors in France (Legrand and Crombez-Bequet, Reference Legrand and Crombez-Bequet2021) found that physical activity significantly improved their self-esteem, which is in line with the experiences reported by the participants of this study who understood physical activity as a protective factor. While focusing on one’s career has not explicitly been explored as a protective factor in IPV, Gibson-Davies and colleagues (Reference Gibson-Davis, Magnuson, Gennetian and Duncan2005) had suggested that empowering women through employment creates the means and support networks needed for independence. This would help address one of the perceived maintaining factors (code 4.2) for participants in this study as many of them did not have the financial means to leave the perpetrator.
The interaction between cognitions, affect, and maintaining factors
There is evidence that IPV that includes isolation, denigration, and humiliation, as is the case of CC, can lead to survivors developing pervasive beliefs about emotional deprivation (‘I am alone’), abuse (‘I deserve the abuse’) and personal failure (‘I am not worthy’), affecting their choice of coping strategy and further maintaining distress (Calvete et al., Reference Calvete, Corral and Estévez2007). These themes echo participants’ experiences, which they linked to increased use of avoidance, further perpetuating psychological distress and concurrent depression.
Participants in this study reported anger as being instrumental to the realisation that they were in an abusive relationship. This is supported by Shurman and Rodriguez’s (Reference Shurman and Rodriguez2006) qualitative interview study with 85 women survivors of IPV, which found that experiencing anger was key to awareness of the abuse. The study also found that increased anger and the internalising emotions like sadness, fear and shame predicted survivors’ readiness to change, and influenced their decision to leave the relationship, further supporting the connection between emotion and behaviour found in this study.
A maintaining factor that has been found previously for IPV survivors staying in abusive situations is social norms around ‘broken marriages’ and non-violent abuse not being considered as damaging as violent abuse (Pokharel et al., Reference Pokharel, Hegadoren and Papathanassoglou2020). These could have interacted with the existing shame that survivors felt for being in the relationship, and prevented them from seeking support.
Unhelpful coping strategies
Traditionally, there has been an understanding that women remain in abusive situations due to denial (Debono, Reference Debono2011; Women’s Aid, n.d.), which was interestingly not found as a coping mechanism in this study. This is in concordance with a study (Foster et al., Reference Foster, Becho, Burge, Talamantes, Ferrer, Wood and Katerndahl2015) conducted with 400 women IPV survivors in the USA, which reported that women are neither in denial nor accepting of the violence in their lives, but that, as reported by the participants in this study (code 2.5 ‘it might get better’), they have some hope that “something dramatic will change”. Substance misuse (code 5.1.4) has also been found commonly amongst survivors of IPV across studies, who are at risk for transitioning to substance use disorder (Mehr et al., Reference Mehr, Bennett, Price, de Souza, Buckman, Wilde and Esopenko2023). Additionally, substance misuse is highly co-morbid with mental health diagnoses, increasing survivors’ victimisation (Mehr et al., Reference Mehr, Bennett, Price, de Souza, Buckman, Wilde and Esopenko2023).
Mental health during and after the relationship
This study is the first to qualitatively explore women’s mental health experiences as an impact of CC. The findings highlight survivors’ experiences of pervasive sadness, depression, and suicidal thoughts during the relationship, and mental health implications in the long term, which include experiences of anxiety, panic attacks, depression, trauma, and chronic mistrust. Lohmann and colleagues’ (Reference Lohmann, Cowlishaw, Ney, O’Donnell and Felmingham2023) meta-analysis on the mental health impacts of CC also found it to be moderately associated with both depression and trauma, comparable to the relationship between psychological IPV and these conditions.
While this study is the first to qualitatively discuss anxiety as a mental health outcome of experiencing CC, a Danish literature review of 10 studies found that women exposed to CC are at a greater risk of experiencing anxiety, depression, PTSD, and suicidal thoughts compared with women who have not been exposed to CC (Frand-Madsen and Lohman, Reference Frand-Madsen and Lohman2023). Anxiety has also long been established in literature as an outcome of general IPV (Chandan et al., Reference Chandan, Thomas, Bradbury-Jones, Russell, Bandyopadhyay, Nirantharakumar and Taylor2020).
Additionally, participants in this study also experienced chronic mistrust of individuals and institutions following their CC relationship and while chronic mistrust has not been explored in general IPV literature yet, existing schemas of mistrust have been found to moderately correlate with IPV victimisation (Pilkington et al., Reference Pilkington, Noonan, May, Younan and Holt2021). The authors argue that individuals with previous experiences related to a lack of safety in relationships may select or tolerate abusive or untrustworthy partners. While the participants of this study did not make this link, the majority of them experienced difficult childhood and previous romantic experiences, perhaps contributing to latent chronic mistrust, which may have been exacerbated after the CC experience. Thus, chronic mistrust could be a maintaining factor in experiences of CC victimisation and should be looked into in future studies.
Women experiencing chronic mistrust of individuals and institutions may also be more vulnerable to isolation even after the end of the relationship, as they experience difficulty in opening up or trusting information from new people (Lieberz et al., Reference Lieberz, Shamay-Tsoory, Saporta, Esser, Kuskova, Stoffel-Wagner and Scheele2021). Additionally, these women may have difficulty trusting institutions, such as the police force, GPs, etc., due to their trust being abused in their intimate relationships. This may mean they do not seek help from public institutions in the future in times of need or crisis (Wolf et al., Reference Wolf, Ly, Hobart and Kernic2003), increasing their risk for further victimisation.
Relevance of the findings to existing psychological models
The findings of this study align with existing trauma models, including PTSD and complex PTSD frameworks, which highlight the profound psychological impact of sustained abuse. Many participants described experiences of emotional ‘shutdown’ and dissociation as coping mechanisms within their CC relationships. This aligns with Ehlers and colleagues’ (Reference Ehlers, Maercker and Boos2000) concept of ‘mental defeat’, which describes a state where individuals feel powerless and detached from their sense of self, commonly observed in PTSD sufferers. The notion of ‘checking-out’ or ‘switching-off’ reported by participants could be conceptualised within this framework, suggesting that CC may lead to similar cognitive and emotional disruptions seen in PTSD.
Additionally, Young’s Schema Therapy (Young et al., Reference Young, Klosko and Weishaar2006) provides a useful lens for understanding the development of vulnerability factors observed in survivors of CC. Many participants described childhood adversity and prior negative romantic relationships, which may contribute to the formation of early maladaptive schemas such as mistrust/abuse, defectiveness/shame, and emotional deprivation. These schemas could contribute to survivors’ difficulties in recognising coercive control and breaking free from abusive dynamics, reinforcing cycles of victimisation. Integrating schema therapy perspectives with this model may offer clinicians additional tools to help survivors restructure deep-seated beliefs and relational patterns. Future research could explore these links further.
Moreover, dialectical behaviour therapy (DBT; Linehan, Reference Linehan1993) may hold relevance in supporting survivors of CC. DBT’s emphasis on distress tolerance, emotional regulation, and interpersonal effectiveness could be particularly beneficial given the emotional dysregulation and difficulties in self-advocacy reported by victims of CC and domestic abuse more broadly. Many survivors in this study described chronic mistrust and heightened fear responses, suggesting that interventions aimed at strengthening emotional resilience and interpersonal skills may complement CBT-based approaches. Future research could explore how DBT-informed interventions may assist survivors in developing more adaptive coping mechanisms beyond traditional CBT-based strategies.
Furthermore, the acceptance and commitment therapy (ACT; Hayes et al., Reference Hayes, Strosahl and Wilson2011) model offers an additional theoretical perspective that may be beneficial for survivors of coercive control. ACT focuses on psychological flexibility – the ability to remain open to experiences, stay present, and take committed action based on one’s values. Many participants in this study described difficulties in trusting their own judgement and feeling trapped in patterns of avoidance, self-doubt, and emotional suppression. ACT’s emphasis on diffusion from unhelpful thoughts, mindfulness, and value-based action may provide survivors with alternative ways of engaging with their internal experiences. By helping survivors shift from avoidance-based coping mechanisms (such as dissociation and emotional numbing) towards an acceptance-based approach, ACT can enable them to reconnect with their sense of self and personal agency, something that has been widely documented within the CC literature as being significantly impacted by the abuse (Moulding et al., Reference Moulding, Franzway, Wendt, Zufferey and Chung2021; Pitman, Reference Pitman2017). Future research could explore how ACT-informed interventions may complement cognitive behavioural approaches by fostering psychological resilience and promoting long-term emotional well-being in survivors of CC.
Strengths and limitations
This study reports on a novel approach to better understand the experiences of victims of CC. The study closely followed grounded theory methodology, employing theoretical sampling, constant comparison, and iterative data analysis, with a sufficiently large and heterogeneous sample to achieve data saturation (Hennink and Kaiser, Reference Hennink and Kaiser2022). The study has involved experts by experience, and the findings have been member-checked by seven participants. Another strength was the involvement of a second reviewer in analysing a sample of transcripts, adding credibility to the findings. Limitations include potential self-selection bias, a white British majority in the sample, retrospective reporting of experiences and exclusion of those with significant concurrent mental health difficulties or safeguarding concerns, limiting the transferability of the results. As stated, the findings from this study provide a preliminary theory of CC experiences, rather than a fully developed and universally applicable model, which will require further refinement and testing.
Implications and future research
While disorder-specific models like trauma-focused CBT are effective for addressing symptoms of PTSD, depression, and anxiety, they may not fully account for the psychological impact of long-term, subtle forms of abuse, such as gaslighting, that are central to experiences of coercive control. This model of coercive control experiences brings these overlooked dynamics into focus, helping clinicians better understand the complex, cumulative effects of sustained psychological abuse, which may not be directly addressed by standard PTSD or depression-focused CBT interventions. By drawing attention to the pervasive and insidious nature of coercive control, this model can guide therapeutic interventions in a more holistic way, providing a clearer framework for understanding the long-term psychological effects that might otherwise go unnoticed.
Additionally, this model opens avenues for further research into the specific impact of CC experiences on mental health. Areas such as the long-term effects of gaslighting, emotional manipulation, and isolation could be further explored, contributing to the development of more tailored therapeutic approaches for survivors. Future research could investigate how the model could be incorporated into clinical practice to enhance existing treatments, as well as how it might inform new interventions specifically addressing the residual impacts of CC on victims. In summary, this study contributes to a more nuanced and comprehensive understanding of the psychological impact of coercive control and provides a foundation for future research to test and refine this integrative framework.
Unlike earlier frameworks that position CC primarily within legal, sociological or perpetrator-focused paradigms, the present model delineates survivor-centred psychological processes and their mutual reinforcement, offering a bridge between CC research literature and evidence-based CBT practice. Psychological professionals can use this initial CBT model to formulate the experiences of their clients, integrating other relevant models as referenced earlier to formulate idiosyncratic formulations. Cognitions such as ‘I am not worthy’, ‘I am alone’, ‘I deserve the abuse’, and ‘It might get better’ may be targeted through cognitive restructuring techniques (Andersson et al., Reference Andersson, Olsson, Ringsgård, Sandgren, Viklund, Andersson and Bohman2021; Perangin-Angin et al., Reference Perangin-Angin, Wijono and Hunga2021); whilst addressing schema-driven relational patterns and equipping survivors with DBT-based emotion regulation strategies may help improve long-term psychological resilience.
This model also highlights the importance of addressing chronic mistrust, as it has significant implications for survivors’ ability to form new relationships and seek institutional support, something that has been found to be a barrier to help-seeking for minoritised groups (NHS Race and Health Observatory, 2025). Additionally, a focus on protective factors, including social support, physical activity, and vocational engagement, may enhance recovery and empower survivors to rebuild a sense of identity beyond the abusive relationship.
Policies and services should explicitly address the complexity and long-term impacts of CC, beyond the immediate risk and crisis management, support early identification, and develop community-based resources. For survivors who are mothers, interventions should consider the additional caregiving barriers they face, with family-based supports developed accordingly. In line with NICE guidelines (National Institute of Health and Care Excellence, 2014) on multi-agency working for domestic abuse cases, professionals working with survivors should be provided training on psychological models of CC.
Future studies should test vulnerability and protective factors quantitatively, examine gender and cultural variations, and evaluate clinical utility.
Conclusion
This study offers a CBT framework for understanding women’s experiences of CC relationships, supporting psychological formulation and intervention planning. The findings highlight women’s perceptions of life events that made them vulnerable to CC relationships, cognitions experienced by survivors and related emotions and behaviours, as well as protective factors. The model offers avenues for treatment that haVE the potential to better address the psychological impact of CC. Replicating and testing the current model in future studies will be helpful in building the empirical evidence needed to develop a CBT intervention for survivors in the future that addresses the key aspects of their unique experience.
Key practice points
-
(1) Coercive control is a frequent form of IPV that is associated with mental health difficulties, and psychological professionals should be provided training on coercive control.
-
(2) Psychological professionals could use this preliminary model to formulate the cognitive, affective, and behavioural experiences of their clients with experience of coercive control. The client’s specific vulnerability factors and the impact of the coercive control experience on their current lives should also be explored.
-
(3) Intervention plans could include prioritising cognitive-restructuring, exploring unhelpful coping mechanisms, engaging with protective factors, and rebuilding interpersonal trust in relationships.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1754470X25100317
Data availability statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to risk of compromising participant privacy.
Acknowledgements
None.
Author contributions
Arshia Amin Choudhury: Conceptualization (Lead), Data curation (Lead), Formal analysis (Lead), Methodology (Lead), Project administration (Lead), Writing - original draft (Lead), Writing - review & editing (Lead);
Olga Luzón: Supervision (Lead), Writing - original draft (Supporting), Writing - review & editing (Supporting).
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
There are no competing interests to disclose.
Ethical standards
Ethical approval for the study was obtained from Royal Holloway’s Research Ethics Committee on 3 March 2023. Considerations were based on the code of human research ethics, code of conduct, and ethical principles of psychologists, as set out by the BABCP and BPS (BABCP, 2022; Oates et al., Reference Oates, Carpenter, Fisher, Goodson, Hannah, Kwiatowski and Wainwright2021). All participants were provided with detailed information about the rationale for the study, their involvement, rights, potential harms, and confidentiality procedures. All participants gave their informed consent to take part.
Open access
Comments
No Comments have been published for this article.