Background

Intimate partner violence (IPV) continues to be a global crisis against women, children exposed to it, and the mother–child relationship. IPV is characterized by the actions of a present or former intimate partner or spouse, leading to physical, sexual, or psychological harm. It stands as a prevalent manifestation of gender-based violence against women on a global scale. Statistics reveal that 27% of women who have been in an intimate relationship have encountered instances of physical and/or sexual violence perpetrated by their partners (World Health Organization, 2024b). During the COVID-19 pandemic, nearly one in two women reported that they, or a woman they know, had experienced violence (UN Women, 2022, p.14–15). In Hong Kong, there were around 2,000 to 3,000 newly reported IPV cases annually in the past five years, and persistently over 80% of the victims were women (Social Welfare Department, 2024). However, the newly reported cases might only represent 2% of actual IPV cases in Hong Kong (Lau & Blundy, 2016), as most of the abused women did not seek help at all or only sought informal support (Chan, 2022). IPV creates immediate death risk and causes negative physical, psychological and mental consequences to women survivors, damaging their economic, political, and social participation (e.g., Crowne et al., 2011; Dillon et al., 2013). IPV is violence against children exposed to it, subjecting them to its detrimental effects, and globally one in four children resides with a mother who is a victim of IPV (Stoltenborgh et al., 2012; Unicef, 2017; World Health Organization, 2024a). Compared with their peers without IPV exposure, these children had a greater risk of developing trauma symptoms, and emotional and behavioral problems that could continue to put them at risk as adults, indicating an intergenerational transmission of impacts of IPV (Adams et al., 2019; Kimball, 2016; Pinheiro, 2006). Although the children might have very close relationships with their mothers who were their sources of comfort and support in the context of IPV (e.g., Lapierre et al., 2018; Mullender et al., 2002) and the mother might attempt to compensate for their children’s exposure to IPV through positive and nurturing mothering (e.g., Letourneau et al., 2007; Pels et al., 2015), IPV is violence against the mother–child relationship. Humphreys et al. (2006) presented a list of tactics perpetrators applied purposely to attack this relationship, abusing the mothers in the presence of the children to show that the mothers were incapable of protecting themselves and parenting the children. In the post-separation stage, the perpetrators could be more present than absent (Humphreys et al., 2019). They might exercise abuse, coercion and control over women through custody, parenting, and visitation contacts (Hayes, 2012; Katz, 2015b; Thiara & Humphreys, 2017).

An ensuring body of research has demonstrated the significance of the mother–child relationship in the context of IPV. Restrengthening this relationship is crucial for the mother–child survivors to support and promote each other’s recovery from the damages of IPV at the post-separation stage (Chanmugam, 2014; Humphreys et al., 2006, 2015; Katz, 2015a, b; Kong & Hooper, 2018; Mullender et al., 2002; Swanston et al., 2014; Visser et al., 2016). The mother–child relationship consists of explicit and implicit aspects that coincide in our communications and interactions, and therefore, are equally important. The explicit relationship includes conscious communications that can be verbalized through spoken language, and the implicit relationship involves nonconscious interactions that can be expressed through nonverbal manifestations, such as art (Gavron, 2013, 2019; Gavron & Mayseless, 2015, 2018). The previous studies mainly focused on the participants’ verbal reports of their explicit relationship. The equally important implicit relationship remained unexplored (Wong & Ho, 2022). Despite the body of research, there is a lack of theory to explain mother–child relationship in the context of IPV. Little attention was given to changes in the mother–child relationship led by changes in the context, such as before and after leaving the perpetrators/fathers. Moreover, the studies were primarily conducted in Western countries, and Chinese mother–child survivors were seldom involved as participants. To the best of our knowledge, this is the first study to involve the Chinese mother–child survivors and explore their explicit and implicit relationships to develop a model explaining the changes in their relationships led by the changes in the context of IPV.

Ethical Considerations

Priority was given to ethical considerations as IPV research is sensitive (Mullender et al., 2002; Van Gelder et al., 2020). Particularly in Chinese culture, IPV is considered a family secret not to be disclosed to outsiders (Sullivan, 2005; Tonsing, 2016; Yick, 2007). This study was framed by the strengths perspective, feminist perspective and child-centered perspective, which were demonstrated to be safe, therapeutic and empowering to women survivors of and children exposed to IPV in the previous studies. The strengths perspective moves away from participants’ problems towards their agency and strengths, thus asserting their perspectives and avoiding labeling them as deficit victims (Mariscal, 2012; Saleebey, 2013; Song & Shih, 2010). The feminist perspective appraises IPV as a gender-based social problem and places the emphasis on the abused mothers’ predicaments, thus avoiding victim- and mother-blaming (Buchanan & Wendt, 2018; Lapierre, 2010; Moulding et al., 2015; Peled & Gil, 2011; Semaan et al., 2013). The child-centered perspective underlines children’s agency and strengths, thus avoiding labeling them as passive witnesses of parental IPV (Katz, 2015a; Kong & Hooper, 2018; Mullender et al., 2002), tackling the unequal adult–child power relation (Dedding et al., 2015; Graham et al., 2015; Haring et al., 2019; Kirk, 2007) between child-and-researcher and child-and-mother, and asserting their interpretations and perspectives in this research context (Wong & Ho, 2022). Ethical approval was obtained from the Human Research Ethics Committee of The University of Hong Kong before data collection. Signed informed consents were obtained from the mothers for their own and their children’s participation. The children’s signed informed consents were sought to show them equal respect, engage their exercise of agency, and ensure their voluntary participation. The participants were allowed to choose the venue for data collection for safety reasons, and they all chose the shelter where they felt safe and supported. Therefore, this study was conducted in the shelter. All the participants received a HK$20 supermarket or bookstore voucher, subject to their options, as a token of thanks for participating.

Method

This study adopted a qualitative research design because it allows the mother–child survivors to inform and interpret their lived experiences in their own words and personal ways (Nikupeteri & Laitinen, 2015; Skinner et al., 2013). Grounded Theory qualitative methodology (Glaser, 1978; Glaser & Strauss, 1968) was applied to conduct individual interviews and mother–child Joint Painting Procedure (JPP) with the participants to explore their explicit and implicit relationships, respectively. Further, it was applied to assert participant perspective as most important (Glaser, 2007), preventing the researcher from stepping inside the mother–child survivors’ expressions and interpretations. Theoretical sampling, data collection and data analysis were carried out iteratively to discover codes, concepts, categories, and the connections between different categories to generate a theory (Glaser, 1978, 2002; Glaser & Strauss, 1968) explaining the changes in mother–child relationship led by the changes in the context of IPV.

Sampling

This study began with purposive sampling, relying on the researcher’s judgment to select the appropriate (Babbie, 2021) and knowledgeable participants (Glaser, 1978, p. 45). A mother-daughter dyad was recruited through the researcher’s (first author) direct contact. This dyad had survived IPV together and had been living in their new home for two years after being discharged from the shelter. The researcher met them at the Welfare Panel Subcommittee Meeting of Hong Kong Legislative Council in 2015, where the mother spoke on behalf of the IPV women survivors’ self-help group. She expressed their needs and gave recommendations for policy and service improvements. When the researcher invited the dyad to participate in this study in 2018, they gladly accepted it and chose to have individual interviews at home. After analyzing the data collected from the dyad, theoretical sampling was applied. Theoretical sampling is the process whereby the researcher collects and analyzes the data collected and then decides what data to be collected next and where to find them to develop a theory emerging from data (Glaser & Strauss, 1968, p. 45). Both the mother and daughter reported in their individual interviews that there were changes in their relationship after leaving the perpetrator/father. Furthermore, they had approached different relational and community resources before entering the shelter as their last resort. This information directed the researcher to recruit Chinese mother–child survivors from a shelter for abused women and their children. Finally, 33 participants (13 mothers and 20 children) were referred by their shelter social workers to participate in this study. This sample size was determined by theoretical saturation, that is, no additional data could be developed into a category and no new category emerged from the data (Glaser & Strauss, 1968, p. 61). The participants’ major characteristics can be found in Appendix 1 and 2.

Data Collection

Participants’ demographic data, immediate concerns, and social backgrounds were collected using a background information sheet developed by the researcher.

Data on the explicit mother–child relationship were collected from semi-structured individual interviews conducted in Cantonese or Putonghua, depending on participants’ preferences. The interviews with mothers lasted about 2.5 hours, and children lasted around 30 minutes to one hour. All interviews were audio-recorded and transcribed verbatim by the researcher.

Data on the implicit relationship were collected from mother–child JPP, the dyadic art-based method developed by Gavron (2019) to examine the implicit parent–child relationship, involving nonverbal and nonconscious interactions (Gavron, 2013, 2019; Gavron & Mayseless, 2015, 2018). In this study, the mother–child dyads experienced the 5-step JPP together on the same paper. At the beginning, the researcher emphasized to the dyads that they were partnering in an enjoyable activity, and that they could paint freely throughout the process. The dyads commenced their JPP by penciling out their own personal space. Then they painted inside their own personal space. After that they painted their own frame around their own personal space followed by painting a path from their own frame to the partner’s frame. Finally, they painted together on the rest of the paper, that is, their common area. During the painting process, the researcher took notes of the dyads’ communications and interactions. After painting, the dyads appreciated their co-created artworks (e.g., Figs. 1, 2, and 3) to discuss their artistic elements step-by-step and one by one within each step according to their co-developed post-painting discussion guide. Wong and Ho (2022) showed how the development and application of the post-painting discussion guide promoted the mother–child survivors’ agency and strengths in deciding what and how to present in their post-painting discussions. This avoided misinterpretation, asserted participant perspectives, and especially supported the children to express their own views without imposition by adults, including their mothers and the researcher, and therefore, provided rich data to understand their implicit relationship. Their post-painting discussions were carried out in Cantonese or Putonghua, audio-recorded and transcribed verbatim by the researcher. A total of 18 JPP sessions were conducted with 12 mothers and 18 children. One mother and her two children moved into their new home soon after their interviews. They did not participate in JPP due to COVID-19 and the children’s learning schedules.

Data Analysis

Coding is the core procedure of data analysis. The mothers’ interviews, children’s interviews, and mother–child JPP and post-painting discussions were coded separately, adhering to the Grounded Theory (Glaser, 1978; Glaser & Strauss, 1968). The coding procedure began with open coding of the transcribed verbatim of recordings to generate codes and concepts in every possible way. Next, the related codes and concepts were grouped into a category. When the connections between the categories emerged, they were further grouped into interrelated subcategories and a core category (Appendix 3). Finally, the core category and interrelated subcategories were integrated and conceptualized into the integrated model (Fig. 4).

Rigor and Trustworthiness

The reliability of coding was ensured. Three independent coders with PhDs and research backgrounds were engaged to perform inter-coder agreements (Guest et al., 2012), resulting in 86.34% agreement for mother interviews, 82.93% for child interviews, and 85.33% for mother–child JPP and post-painting discussions. Also, the transcribed verbatim of audio recordings were audited by one of the coders, and no major discrepancy was identified. Another coder examined the English translations of the codes, concepts, and direct quotations originally in Chinese. Discrepancies in translations were resolved in consensus discussions. Moreover, different measures were applied to ensure the trustworthiness of the research process (Lincoln & Guba, 1985). For example, a prolonged engagement that the researcher with over a decade of experience in the field of family violence possesses a strong comprehension of the participants; regular debriefing between the researcher and supervisor (second author) that guaranteed appropriate use of different research methods; triangulation that multi-agents (mothers and children) and multi-methods (individual interview and JPP) were employed to understand the explicit and implicit mother–child relationships from multi-data sources (participant background information sheet, maternal report, children’s self-report, researcher’s fieldnote, mother–child JPP artwork and post-painting discussion). More importantly, member checks were conducted with the participants at three stages (after opening coding, theoretical saturation and conceptualization) to ensure the findings asserted their perspectives. Member checks of individual interviews were conducted individually, and JPP and post-painting discussions with the mother–child dyads. For ease of the participants’ reference and recollection of memories, the findings were organized chronologically to reflect their experiences. Also, the categories were supported by vivo codes and verbatim quotes to directly reflect what they said in their interviews, painting process and post-painting discussions. Based on their feedback, some adjustments were made to the findings to assert their perspectives. For example, the participants underlined that they were each other’s partners (“拍檔” in Cantonese and Putonghua) in both their JPP and daily life. Especially the participants who had settled down in their new homes after being discharged from the shelter related their partnership in JPP to their new home, new life. Therefore, partnership emerged as one of the categories that vividly reflected their expression and choice of words.

Findings

The integrated model, “Dynamic changes in mother–child relationship in the context of IPV” (Fig. 4), is grounded in three main parts of findings: First, the mother–child survivors’ intersecting social identities and backgrounds (Appendix 1 and 2); second, the explicit mother–child relationship (Appendix 3); and third, the implicit mother–child relationship (Appendix 3).

Mother–Child Survivors’ Intersecting Social Identities and Backgrounds

Thirteen mothers and their 20 children participated in this study. The 13 Chinese mothers were between 25 and 45 years old, and their mean age was 35.08 years (SD = 6.21). The major characteristics of these mothers were their intersecting social identities (Appendix 1). All of them were homemakers and primary caregivers of their children regardless of their educational backgrounds. They were living on social welfare with their children, reflecting their poverty. They endured multiple forms of IPV and suffered from physical injuries as well as emotional breakdowns. The perpetrators were their husbands or ex-husbands and the biological fathers of their children, and therefore, were addressed as perpetrators/fathers in this study. Maternal reports reflected that 10 out of the 13 perpetrators/fathers were social welfare recipients and the remaining three were manual workers. The mothers remained in their abusive relationships for 7 to17 years because they wanted to maintain their marriages, hoped the perpetrators/fathers would change, and preferred their children to grow up in intact families. They reasoned their intersecting social identities as women, wives and mothers that deterred them from leaving their abusive relationships. Eleven mothers were migrants from China, and the remaining two were local Hong Kong people. The migrant mothers moved to Hong Kong after obtaining residency, and their household registrations in China were canceled. They did not return to China when IPV occurred and escalated: “He gambled… and beat me up… I hoped to improve our relationship… He refused and continued to gamble…, and the beating became more intolerable… I did not take my daughter back to China… our household registrations were no longer there… I was collapsing!” (Mother, family 9) All the migrant mothers noted they were trapped in their abusive relationships, and 9 of them reported being diagnosed with chronic illnesses due to IPV victimization, e.g., post-traumatic stress disorder (PTSD), depression, and anxiety disorder. All the migrant mothers also informed that they were financially deprived, lacked kinship and social supports, were unaware of community resources, and did not have the required qualifications or skills to get a job in Hong Kong. Thus, relying on the perpetrators/fathers was what they thought best for their own and their children’s survival before admission to the shelter. Twelve mothers were the first time leaving the perpetrators/fathers and were admitted to the shelter together with their 19 children,¹ and the remaining one mother and her child the second time. None of the mothers indicated they would return to the perpetrators/fathers.

Owing to their mothers’ intersecting social identities, the major characteristics of the children were their intersecting social backgrounds. The 20 children (Appendix 2) were between 7 and 13 years old; their mean age was 8.95 years (SD = 1.53). All the children reported financial poverty in their families and attributed it to the perpetrators’/fathers’ problems, for example, addictive gambling and alcoholism. All of them eye-witnessed their fathers’ escalating violence and their mothers’ IPV victimization. All the children conveyed that they were unable to stop their fathers’ violence or change their family circumstances due to their young age: “My younger brother, sister and I were in our room… We heard that Mama was crying and yelling… We were small children… We locked ourselves in our room… because we were too afraid of Baba!” (Elder daughter, family 8) Eighteen of them were born in Hong Kong and the remaining two were in China. Further, 14 of them were brought up by their mothers in China, whereas their fathers stayed in Hong Kong. When they reached school age, their mothers took them to Hong Kong for education and family reunions.

Explicit and Implicit Mother–Child Relationships

The core category, mother and child valuing each other throughout the process consisted of three interrelated subcategories: First, transitions in the context of IPV; second, explicit relational dynamics; and third, implicit relational dynamics (Appendix 3).

Transitions in the Context of Intimate Partner Violence

The mother–child survivors informed that they journeyed through transitions in the context of IPV from the original home with IPV to the shelter and then the new home, new life. They also projected a better future.

Original Home with Intimate Partner Violence

All mother–child survivors reported the perpetrators’/fathers’ escalating violence associated with other problems, for example, alcoholism and substance abuse: “Initially, Baba scolded Mammy in the midnight… I knew that they tried to cover up… Baba lost money in gambling… Then I saw him hit Mammy… Before coming here (shelter), Baba was too violent! No one could stop him! He pulled Mammy’s hair, knocked her head against the wall… punched her… kicked her… Mammy couldn’t tolerate anymore.” (Daughter, family 9) The mothers recalled their endurance of IPV until the hit bottom incident, which they defined as when IPV escalated to become particularly dangerous and intolerable, pushing their emergent separation from the perpetrators/fathers. They stated their commitment to being together with their children, otherwise, they chose to stay in their abusive relationship until they found an opportunity to escape together with their children: “I was no longer able to tolerate… I stayed in the relationship… as I wouldn’t leave my children behind and run away myself… The opportunity came when I took my children to the town hall for my daughter’s ballet examination… I reported to the police and told them that I was afraid of returning home… afraid of his violence… The police asked us to go to the shelter straightaway.” (Mother, family 1).

Shelter

All the mother–child survivors appreciated that the shelter was safe with professional support, regardless of some inconveniences, such as short-term and group living: “I was so worried after leaving him (perpetrator/father). My social worker here is supportive… I’m a new immigrant mother with two children… The social worker helped us obtain welfare services… Here is temporary, we can’t stay for long… the social worker helped us with matters related to housing and money… Now, I have learned that so many places and people are willing to help us in Hong Kong… I felt safer… when I realized that I needed not to return to him.” (Mother, family 1) “The social workers here are very nice to us… I have good friends to play with… I asked them to remember to call me after our discharge.” (Daughter, family 12).

New Home, New Life

The participants who were ex-residents of the shelter informed that they obtained support from various sources to settle down in their new home, new life: “Our family social worker referred us to a lot of resources…to set up our new home… The school social worker continues to help my daughter… to live our new life together.” (Mother, family 12). The mothers and children consistently described that their new households were in-charged by the mothers who devoted to providing a stable environment for their children’s best development: “In the past, my children were very quiet and cried easily in their schools due to our family matters… We have settled down in our new home… I do my best… We can live our own lives in Hong Kong… We don’t need him (perpetrator/father)!” (Mother, family 3) The children noted that their schooling became more stable than before. They spent most of their energy and time with their peers in school. They also obtained support from their schools due to their family background with IPV: “When staying with Baba, I couldn’t sleep… I was late to school… slept in the class… Now, our new home without Baba… I don’t need to worry about Mammy being hurt by Baba when attending class in school… the school social worker talks to me… my class teacher encourages me… I want to do better in my English.” (Son, family 2).

Better Future

All the mother–child survivors foresaw that they would overcome their existing difficulties. They projected their better future consisting of improved family circumstances, independence and mutual care: “Now, we are safe… Mammy doesn’t need to be afraid of Baba… We will be happier… When I grow up, I want to join the Hong Kong Police… to protect Mammy.” (Son, family 13).

Explicit Relational Dynamics

Experiencing the transitions in the context of IPV together, their relationship manifested a combination of explicit relational dynamics: Interdependence, boundaries, and partnership, which co-existed in their explicit relationship, and a particular one appeared more salient than the others in a certain transition.

Interdependence

The mother–child survivors explicitly expressed that they valued each other, and their relationships had always been good and intimate. They emphasized their interdependence and the prime importance of being together throughout their experiences. The mothers conveyed that the children were their meaning in life. The children expressed that they were happiest being together with their mothers. Both mothers and children described their interdependence appeared more salient in their original home with IPV than in other transitions. All of them reported their common difficulty in being unable to stop the perpetrators’/fathers’ violence in their original homes: “I saw Baba hit Mammy repeatedly… He was drunk… no one could stop him… Mammy was in despair.” (Daughter, family 10) Their common concerns were each other’s safety and their family being intact. To ensure safety, they became each other’s protectors and advisors, and allied to resist the perpetrators’/fathers’ escalating violence and other problems: “He became crazy at midnight… I purposely yelled to wake up my children… My children knew my difficulty… allied with me… yelled and cried loudly together, and this could restrain him a bit… otherwise I would have been killed by him long ago!” (Mother, family 2). At the same time, to maintain an intact family, they allied to keep the perpetrators’/fathers’ violence and other problems family secrets. They did not disclose what they were going through to outsiders, restraining them from seeking external support. They became each other’s best friend to meet their emotional needs: “To preserve his image, I didn’t tell my own families about his violence and drug problems… I don’t want them to look down on him… My son saw him beating me… tried to protect me… comfort me… We talk about everything… We feel the same way… We are best friends… confidants.” (Mother, family 7) The migrant mothers emphasized: “My two children are my only family ties in Hong Kong… They are my companions… my confidants… Three of us depend on each other!” (Mother, family 3).

Boundaries

The mothers and children consistently informed that there were boundaries between them regardless of their interdependence. Their boundaries were more salient in the shelter than in other transitions. The children conveyed their need for more intimacy and reassurance from their mothers as the shelter was a new environment. On the contrary, the mothers explicitly expressed their need for personal space away from their children to face the challenges after leaving the perpetrators/fathers and to manage their emotions: “After coming here, my son wanted me to be with him more… I tried to keep a distance from him…I needed to settle my emotions… I was afraid that I might lose my temper and vent on him.” (Mother, family 7) After having adjusted to the shelter, 10 out of the 13 mothers considered divorce (the remaining three mothers had divorced before their admission to the shelter). However, 16 out of the 20 children preferred parental reconciliation over divorce: “The best solution is for Baba and Mammy to reconcile… Baba should stop drinking… and he should stop hitting and kicking Mammy.” (Daughter, family 1) Knowing their children’s different standpoints, the mothers made clear that divorce was an adult matter under their decision-making domain, and custody was within their children’s: “Divorce was an adult’s matter … between his Baba and me … that’s my decision, my domain… I discussed custody with my son… in case he was asked about his preference of custodial parent.” (Mother, family 7) In the process of managing divorce and related matters, the mothers depended less on their children than before. Instead, they became more assertive and appreciated supports from professionals and other mother survivors of IPV: “Divorce is the biggest struggle… is a taboo in our hometown… I have a lot of worries… my children can’t understand… I talked to my very supportive social worker… I also talked to other mothers of similar situation…” (Mother, family 1) The children noticed their mothers’ changes and wanted to maintain their best-friend relationship with their mothers, but they found that the social workers were more effective in consoling their mothers than themselves: “Now, Mammy talks less to me… She talks more to Ms. W (the shelter social worker). I am sure that Mammy has something in her heart, but she doesn’t tell me… I pamper her to make her happy.” (Daughter, family 9) Despite their different standpoints over parental divorce, all the children appraised their mothers as better parents than their fathers and other caregivers, and therefore, assertively chose their mothers as custodial parents upon parental divorce.

Partnership

All the mothers and children emphasized their partnership throughout their experiences, for example, to maintain their daily family functioning: “We are Mammy’s partners… We take turns to help Mammy with the housework now and before leaving Baba…” (Son, family 8), and to keep shelter location confidential: “Baba asked me about our location… I didn’t let him!” (Son, family 13) In particular, all the ex-resident participants said that their partnership appeared more salient in the new home, new life than in other transitions. To nurture their new families, they collaborated to minimize contact with the perpetrators/fathers when they realized such contacts would induce conflictual dialogues between them: “When we just moved into our new home… My son was very frustrated and bit my hand… He said that he liked Baba and missed Baba… I was so apathetic… Now, my son is much different… Maybe he understood that I devoted so much to our new household.” (Mother, family 3) “I am stressed when making child visitation arrangements with their Baba… I will not let him come to our new home. Never! My children understand my difficulties… They collaborated with me… They ignored him.” (Mother, family 2) The mother–child survivors also performed their individual responsibilities and expressed their strong mutual recognition in their new homes – the mothers took charge of their new households to provide the best for children’s development, and the children focused on their studies: “Mammy has been doing a lot in setting up this new home… Now I want to improve my English… to make Mammy happy!” (Son, family 4) “My children accompanied me… understand my difficulties… are very concerned about me… I am so thankful to have them!” (Mother, family 2).

Implicit Relational Dynamics

The synergy between JPP and the co-developed post-painting discussion guide circumvented the difficulty in verbalizing relational matters and optimized children’s active participation. It facilitated bidirectional and spontaneous mother–child dialogues. Their interpretations of artistic elements represented their “here and now” inner experiences, feelings and thoughts, thus eliciting information on their implicit relationship that could have been otherwise difficult to obtain (Wong & Ho, 2022). Three implicit relational dynamics emerged from the data: Togetherness, boundaries, and partnership, which are illustrated by the following case examples of the mother–child JPP under pseudonyms:

Togetherness

All the mother–child dyads expressed their togetherness in three ways. First, they mutually engaged in their painting process, and many of the children sang happily. Second, they created various artistic elements to represent each other and the importance of being together. Third, they, especially the children, expressed that they did not want their JPP to end and needed more dyadic activities (Wong & Ho, 2022). In Fig. 1, Joyce (mother, 41-year-old) and James (son, 9-year-old) expressed their togetherness when they interpreted their personal space. James said, “The heart means togetherness… I thought of Mama and myself…” Joyce followed, “My sunflower means our long-lasting love… It is most important for us to be together and love each other.” Further, they explained that the blue and red bridge at the lower part of the paper was their common path, and the two ants on it represented each other. They interpreted this as themselves walking through different circumstances together. They appreciated their new co-creation experience and titled their artwork, “Loving Together.” They expressed that they could have more similar activities together in the future, reflecting their need for more dyadic activities. In Fig. 2, Susan (mother, 32-year-old) and Steven (son, 9-year-old) agreed to paint birds in their common area. When Steven was painting his birds, he spontaneously sang the popular Cantonese nursery song, A Bird Falling in the Water. After singing, he conversed playfully with his mother about the birds and the song. They conveyed the importance of being together through interpreting their paths. Susan’s path was formed by the grey horizontal lines surrounded by Steven’s two rows of brown and grey dotted path. Susan said, “Looking at our paths… They are a good match…. That means whatever we do… although there will be some obstacles… we are heading together to the sun, a bright and brilliant future.” Steven further interpreted their paths, “Mammy you see, our paths are beautiful. We can walk from the sun, along our paths, to the Christmas tree and take the presents under it!” Then Steven expressed, “I want to paint again with Mammy!” He asked the researcher for a few more drawing papers to use together with his mother afterwards, expressing his need for more dyadic activity.

Boundaries

The dyads’ interactions in their painting process uncovered the mothers’ needs for personal space boundaries and their children’s abilities to learn to respect their boundaries (Wong & Ho, 2022). All the mother–child dyads showed their abilities to keep their own and respect their partners’ boundaries, especially their personal space boundaries. This allowed them to continue to take pleasure in their JPP. The mothers highlighted their mothering role when they interpreted their paintings. In Fig. 3, when Nina (mother, 44-year-old) painted her pink path to reach Natalie’s (daughter, 10-year-old) black frame, Natalie thickened the concerned part of her black frame. She reminded Nina not to get into her personal space, showing her ability to keep boundaries. Nina did not paint on Natalie’s black frame or inside her personal space. In their post-painting discussion, Natalie appreciated Nina for respecting her boundaries and said, “I am happy that Mammy did not paint on my frame or enter my personal space… The girl inside my personal space is me… I am enjoying myself… I don’t want to be bothered.” When Nina interpreted her painting, she highlighted her mothering role, “It’s me inside my personal space… I am a cheerful Mama… taking care of my two children.” She further emphasized her mothering role when she talked about her most favorite part of their painting, “I like the garden most (in the middle of their painting), I like taking my children to the park… They can play… enjoy the sunshine… It’s very good for their health… I am their mother… I am very concerned about my children’s health and happiness…” Natalie appreciated her mother’s interpretation and said, “Mammy, the garden is beautiful!”.

Partnership

All the mother–child dyads partnered in painting process to accomplish their co-created artworks. They also partnered in post-painting discussions to interpret their artistic elements within each painting step. Moreover, they expressed their mutual recognition of each other’s contribution to their co-created artworks. Furthermore, they reached a consensus in painting their common areas and entitling their artworks. In Fig. 1, Susan and Steven expressed their mutual recognition. Susan appreciated, “A Christmas tree on the seaside… so comfortable together… Perfect match!” Steven responded, “Mammy’s sunshine is beautiful. It’s really shining… as shining as the magic star on the top of my Christmas tree!” In Fig. 2, Joyce and James reached their consensus to divide the paper into two halves so that they could paint freely without disturbing each other. When they discovered no space left for painting their paths, they reached their consensus to paint their common path with two ants on it. After painting, they appreciated their beautiful artwork together and expressed their mutual recognition for each other’s contribution. When they were invited to relate their JPP experience to their daily life, James said, “Mama and me are good partners…” Joyce responded, “This is most important… as we are now striving together for our new life.” In Fig. 3, when Nina and Natalie were invited to relate their JPP experience with their daily life, they appreciated each other’s contribution to achieving their co-creation. They reflected that their partnership in their daily life allowed them to be “Happy Every Day,” which became the title of their artwork.

The Integrated Model

The integrated model (Fig. 4), “Dynamic changes in mother–child relationship in the context of IPV”, is developed by integrating the three main parts of findings: Frist, the participants’ intersecting social identities and backgrounds, suggesting IPV is an intersectional social problem. Second, the salient explicit relational dynamics in the relationship varied in different transitions in the context of IPV, showing a 4-stage process form independence in original home with IPV to boundaries in the shelter, partnership in new home, new life, and independence and mutual care in a better future. Third, the dynamic, bidirectional, and reciprocal nature of the dyadic implicit relational dynamics illustrated that both the mothers and children were active agents influencing their communications and interactions. Both the explicit and implicit mother–child relationship consistently reflected the dynamic and changing nature of the relationship in the context of IPV instead of static.

Stage 1: Interdependence in Original Home with Intimate Partner Violence

The mother–child survivors with intersecting social identities and backgrounds presented the strongest interdependence in the original home with IPV, where their common difficulty was being unable to stop the perpetrators’/fathers’ escalating violence and other problems. Further, their common concerns were each other’s safety and their family being intact. They allied to withstand IPV and to keep it their family secret not to be disclosed to outsiders, thus restraining them from seeking external support. They depended on each other to meet their emotional needs and became each other’s best friends and confidants. Therefore, they were more likely to be emotionally fused, and their boundaries appeared less permeable than in other transitions. The mothers endured IPV until the hit bottom incident, pushing their emergent separation from the perpetrators/fathers and admission to the shelter together with their children.

Stage 2: Boundaries in Shelter

Mother–child boundaries were more evident in the shelter than in other transitions. The shelter was a safe venue with professional support. The mothers explicitly expressed their need for personal space away from their children to deal with their emotions and challenges after leaving their original homes. However, the children needed more intimacy and comfort from their mothers in the new group living environment of the shelter. Having adjusted to the shelter, the mothers considered divorce, although most children preferred parental reconciliation. The mothers declared divorce their decision-making domain outside their children’s scope. In the process of managing divorce and related matters, the mothers appreciated support from professionals and other IPV mother survivors. The children tried to make their mothers happy to maintain their best-friend relationship but found that the social workers were more effective in consoling their mothers than themselves. All children assertively chose their mothers as custodial parents upon parental divorce, as they considered their mothers to be better parents than the perpetrators/fathers.

Stage 3: Partnership in New Home, New Life

The mother–child survivors demonstrated a strengthened partnership in the new home, new life than in other transitions. To nurture their new families, they collaborated to minimize contact with perpetrators/fathers when they realized such contacts would lead to conflictual dialogues between them. Additionally, they performed their individual duties, such as the mothers taking charge of their new households to provide a stable environment for the children and the children concentrating on their studies to make their mothers happy. Moreover, they showed strong mutual recognition, the mothers appreciated their children’s understanding, and the children praised their mothers’ devotion to caring for them and their families. The mother–child survivors obtained support from various sources to overcome their difficulties settling down in their new home, new life. They could divert the tensions through their various community and social outlets when conflictual dialogues occurred between them. Therefore, their boundaries were becoming more flexible than before.

Stage 4: Independence and Mutual Care in a Better Future

The mother–child survivors explicitly expressed the prime importance of being together in the process of journeying the transitions in the context of IPV. Additionally, they obtained more support from various sources throughout the process. These findings helped explain their consistent projection of a better future, one in which their family circumstances would be improved, and they would become more independent and continue their mutual care.

Mother and Child Valuing Each Other throughout the Process

The mother–child survivors consistently showed that they had been valuing each other throughout their IPV experience, JPP and post-painting discussions. In their individual interviews, they explicitly stated the prime importance of being together when journeying through the transitions in the context of IPV. In their JPP and post-painting discussions, they mutually engaged in their bidirectional and reciprocal communications and interactions, which elicited their togetherness, boundaries and partnership.

Discussions

The integrated model captures the Chinese mother–child survivors’ intersecting social identities and backgrounds, and their dynamic relationship in the changing context of IPV. It substantiates the collective influence of the context, mother and child on mother–child relationship (Belsky, 1984; Bronfenbrenner, 1977; Feldman, 2010). Its significant implications for professional interventions at the post-separation stage are discussed as follows:

Intimate Partner Violence is an Intersectional Social Problem

The mother–child survivors’ intersecting social identities and backgrounds showed an intersection of poverty, gender and migrant identity that placed them in multiple interlocking difficulties and shaped their IPV experiences. These findings are associated with the central idea of the concept of intersectionality (Crenshaw, 1991, 2021) that IPV does not affect all women and children equally (Collins, 2000; Etherington & Baker, 2018), highlighting the need for recognizing the diversities among mother–child survivors to render effective professional interventions, and more importantly, the need for advocating social and policy changes to better support migrant mothers and their children living with poverty in the context of IPV. In this study, all the mother–child participants were living on social welfare, and the perpetrators/fathers were either relying on social welfare or manual workers, reflecting their low socioeconomic status (SES). Although IPV occurs in all settings and among all socioeconomic, religious and cultural groups (World Health Organization et al., 2013), research has established that men with lower SES appeared more likely to abuse their wives (Abramsky et al., 2016; Resko, 2010). In Hong Kong, it is found that men who married Mainland women have low SES, which directly affects their ability to acquire resources (Chiu, 2017), and their use of violence against their migrant wives can be a form of compensatory masculinity to reclaim power in their private spheres (Cheung & Chiu, 2019). In addition to their use of violence, all the perpetrators/fathers had other problems, for example, substance abuse and addictive gambling. This may imply the need for bringing together practitioners from diverse professions to generate new ways of working with the intersection of these complex problems (Humphreys et al., 2022). In this study, all the mothers suffered from multiple forms of violence. Feminist literature argues that gender inequality is the leading cause of violence against women rooted in patriarchy, placing women as inferior and subordinate to male domination (Galvani, 2006; Tonsing, 2016). This feminist argument endorses the prevalence of gender stereotype within and outside families in Hong Kong (Equal Opportunities Commission, 2020; Women’s Commission, 2018), thus explaining why over 80% of IPV victims are persistently women in Hong Kong (Social Welfare Department, 2024), and more importantly, substantiating the need for incorporating gender as an explanatory factor of IPV (Sánchez-Prada et al., 2023). In this study, the mothers stayed in their abusive relationships for 7 to 17 years, and the migrant mothers highlighted divorce as their biggest struggle, a taboo in their hometowns. These findings support the dual stigmatizations of being a marriage migrant and a single mother that could heighten the migrant mothers’ fears or hesitation to seek external help (Chiu, 2017). The migrant mothers conveyed that they did not return to China when IPV occurred and escalated. Further, they had a lack of external social support in Hong Kong, and therefore were trapped in their abusive relationships. This increased their interdependence with their children to survive IPV together, suggesting the intense mother–child emotional bonding and interconnected boundaries may be an adaptive coping strategy rather than being a problematic or pathological issue (Chanmugam, 2014; Hooper, 2007) in the context of IPV.

A Process of Dynamic Changes in the Explicit Mother–Child Relationship

Findings of explicit mother–child relationship suggest that it is important for professional interventions to help the mother–child survivors recollect and reflect on the process of change in their explicit relationship led by the changes in the context of IPV. These mother–child survivors recollected their experience of withstanding IPV and its impacts together throughout the transitions in the context of IPV, such as from the original home with IPV to the shelter, and new home, new life. They reflected on their explicit relationship throughout this process and conveyed the changes in the salient explicit relational dynamics, including interdependence, boundaries, and partnership, owing to the changes in the context. More importantly, they unequivocally stated the prime importance of being together, and explicitly expressed that they valued each other, allowing them to strive together for their independence and mutual care in a better future. Their recollections of journeying through different transitions in the context of IPV and reflections on their explicit relationship provided a broad picture and new perspective to understand the dynamic changes in their explicit relationship as a 4-stage process: From interdependence in original home with IPV to boundaries in shelter, partnership in new home, new life, and toward independence and mutual care in a better future. These findings are compatible with previous studies, that abused women’s parenting and their children’s adjustment behaviors could be heavily influenced by the context (Levendosky & Graham-Bermann, 2001). The mothers and children were interconnected and interdependent in surviving IPV (Chanmugam, 2014; Katz, 2015b). When transitioned from their original homes to the shelter, their supportive relationship and professional assistance helped ease their difficulties in facing the changes (Humphreys et al., 2006; Jarvis et al., 2005). They could be partners in promoting each other’s recovery at the post-separation stage of IPV (Humphreys et al., 2015; Katz, 2015b; Kong & Hooper, 2018). They could reconstruct their IPV experiences into an empowering perspective of their future regardless of mother–child boundaries and traditional hierarchical communication (Lai, 2011).

Implicit Relationship: A New Dimension for Professional Interventions

Findings of the implicit mother–child relationship imply that dyadic art-based approach may be effective in overcoming mother–child survivors’ communication challenges and strengthening their relationship at the post-separation stage of IPV. Participants’ freedom to paint and interpret their artistic elements was emphasized in their JPP sessions, as disclosing IPV experiences is distressing and overwhelming (McManus et al., 2013; Swanston et al., 2014; Van Gelder et al., 2020). IPV gives rise to a “conspiracy of silence” between the mother–child survivors, that they conceal their feelings and thoughts to protect each other from the painful experience of IPV (Humphreys et al., 2006; Mullender et al., 2002), impedes children’s verbal language development (Conway et al., 2021), and undermines mother–child comfort dialogues and open communication, especially their expression of emotions and affection (Godbout et al., 2009; Kamody et al., 2020; Visser et al., 2016). These challenges may be intensified among Chinese mother–child survivors for cultural reasons. In Chinese culture, IPV is considered a family secret not to be disclosed to outsiders (Sullivan, 2005; Tonsing, 2016; Yick, 2000, 2007). Mother–child communications are indirect, hierarchical and unbalanced (Kong & Hooper, 2018; Lai, 2011). Scant attention is paid to people’s feelings (Sullivan, 2005). Expression of feelings is considered a sign of weakness and strict control over it is valued (Hwang, 1997). Moreover, Chinese parents express less affections to their children than American parents (Zhang & Wills, 2016). These communication challenges were overcome through the synergy between JPP and the co-developed post-painting discussion guide. It provided a relaxing context for the mother–child dyads to experience new ways of communication and interaction. The mother–child dyads enjoyed the painting process. In their post-painting discussions, they visualized their co-created artworks to interpret their artistic elements and listen directly to their partners’ interpretations. They appeared to be ready to communicate with each other through interpreting the meanings of their artistic elements and contents of paintings. They showed interest in each other’s interpretations and paintings, appreciated them, and asked follow-up questions. These allowed them to correctly understand each other’s “here and now” feelings and thoughts without imposition (Wong & Ho, 2022). Their bidirectional and spontaneous communications and interactions revealed that they valued their togetherness, respected each other’s boundaries and achieved their strong partnership. These findings provide a new dimension of evidence to substantiate previous studies conveying that both the mothers and children were active agents contributing to their relationship (Goldblatt et al., 2014; Katz, 2015a; Kong & Hooper, 2018; Neustifter et al., 2015). The mothers could function well in parenting regardless of their IPV victimization (Letourneau et al., 2007; Pels et al., 2015; Tailor et al., 2015). Further, the children appreciated their mothers as their primary sources of support and protection in the context of IPV (Aymer, 2008; Katz, 2015b; Lapierre et al., 2018).

Limitations

This study represents a preliminary attempt to explore the implicit mother–child relationship in the context of IPV using an art-based method, JPP, with Chinese mother–child survivors. Due to the researcher’s background in social work, the data analysis mainly focused on the participants’ communications and interactions in the process and their interpretations of their own paintings. Although findings provide initial information on their implicit relationship, more studies are needed to identify the meanings in their artworks about the mother–child relationship in light of IPV experiences. Further, aesthetic analysis of their artworks is recommended as it may provide additional information to better understand the relationship. Moreover, it should be noted that all the Chinese mother–child participants received shelter service and consented to participate in this study together. Twelve out of the 13 participating mothers were the first time leaving the perpetrators/fathers and the remaining one the second time. None of the mothers indicated that they would return to the perpetrators/fathers. All the children assertively chose their mothers as custodial parents upon parental divorce and collaborated with their mothers to live their new life together. These Chinese mother–child survivors experienced together the 4-stage process one by one. Therefore, their perspectives might provide a limited generalization to mother–child survivors from other cultures or ethnic groups, those who did not receive shelter service together or did not consent to participate together in this study, and those who repeatedly return to the perpetrators/fathers after separation and experienced non-sequential recovery journey (Carman et al., 2023).

Conclusion

The integrated model asserts the Chinese mother–child survivors’ perspectives and contributes a broad picture, new perspective and new dimension to better understand the mother–child relationship in the context of IPV. It carries significant implications for professional interventions at the post-separation stage. First, the Chinese mother–child survivors’ IPV experiences were shaped by poverty, gender inequality, and migrant identity, suggesting the needs for recognizing diversities among mother–child survivors, and advocating social and policy changes. Second, the mother–child survivors unequivocally stated that they valued each other and the prime importance of being together when journeying through the transitions in the context of IPV. This allowed them to strive together for their independence and mutual care in a better future. These findings suggest that professional interventions help them realize the process of dynamic changes in their explicit relationship led by the changes in the context of IPV. Moreover, their mutual engagement in their JPP and post-painting discussions allowed them to overcome their communication challenges. They could express their own and comprehend their partners’ “here and now” feelings and thoughts, implying dyadic art-based approach may be effective for strengthening the mother–child relationship at the post-separation stage of IPV.

Appendices

Appendix 1

Appendix 2

Appendix 3