According to estimates from 2023, approximately 86,200 children under the age of 18 had been registered as having disabilities in South Korea, reflecting a 20% increase from the nearly 72,100 children registered in 2016 [1]. This figure accounts for approximately 1.2% of the total child population, indicating a corresponding increase in the number of families raising children with disabilities. In the process of caregiving, these families face psychological, physical, and financial difficulties, in addition to challenges in social relationships. Among these difficulties, studies find that psychological distress is the most critical [2].

Although both parents experience psychological difficulties, mothers who typically serve as the primary caregiver experience significantly higher levels of parenting stress than do fathers [3]. They often attribute their child's disability to themselves, experiencing guilt, self-blame, and social isolation [4]. Over time, excessive identification with their child can weaken mothers' sense of self and intensify psychological distress. In severe cases, this may lead to suicidal ideation and threaten the stability of the entire family system [5]. Despite these difficulties, mothers have also reported meaningful and positive experiences during the parenting process. Some mothers attach positive meaning to their child's disability, regarding it as an opportunity to strengthen family bonds, develop empathy for others' pain, and adopt a more optimistic outlook on life [4]. By separating their identity from that of their child and overcoming excessive identification, mothers can recognize both their own and their child's inherent worth, which facilitates harmonious and autonomous relationships [5]. Mothers with higher levels of self-compassion and a detached attitude toward difficult situations are more likely to view life positively and demonstrate higher psychological well-being [6]. Moreover, mothers reporting greater parenting self-efficacy tend to practice positive parenting behaviors based on acceptance and understanding of their child, subsequently improving life satisfaction and subjective happiness [7]. Accordingly, self-compassion and parenting self-efficacy can be considered key variables for promoting the mental health of mothers raising children with disabilities, highlighting the importance of assessing such variables.

Self-compassion refers to accepting one's pain with kindness and care, comprising self-kindness, common humanity, and mindfulness. It fosters positive self-awareness that reduces self-judgment, isolation, and overidentification, thereby alleviating depression and anxiety while promoting emotional stability [8]. Neff and Faso [9] found that self-compassion alleviates depression stemming from shame and self-blame in mothers of children with disabilities. In particular, it helps mothers soothe themselves when confronting their children's emotional outbursts or compulsive behaviors, enhancing emotional resilience and improving parenting outcomes. Furthermore, self-compassion facilitates emotional regulation, supports emotional stability, and reduces caregiving burden while enhancing parenting self-efficacy [10].

Parenting self-efficacy, which is derived from Bandura's concept of self-efficacy, refers to parents' belief in their ability to raise and discipline their children and effectively address problems encountered during the parenting process [11]. Park et al. [7] emphasized that confidence and conviction in parenting abilities are critical in increasing life satisfaction and subjective happiness among mothers of children with disabilities. Furthermore, mothers with high parenting self-efficacy were less likely to rely on passive or pessimistic coping strategies when facing their children's problematic behaviors. Instead, they demonstrate active coping and maintain a positive outlook, leading to exceptional parenting practices [12]. Therefore, interventions that enhance self-compassion and parenting self-efficacy are needed to improve the mental health of mothers raising children with disabilities and support effective parenting.

Previous domestic intervention studies targeting these mothers have primarily utilized art therapy [13]. However, more recent research trends have shown an increase in interventions based on theoretical approaches, such as positive psychology and acceptance and commitment therapy [14]. Single-approach psychological interventions include solution-focused group counseling programs that improve self-esteem, parenting self-efficacy, and social problem-solving skills [15] and logotherapy-based programs that enhance life meaning, self-respect, and parenting self-efficacy [16]. These studies suggest that approaches aimed at improving strengths and positive awareness were more effective in improving psychological resilience than those focused solely on pathology. In contrast, although several studies have investigated the effects of interventions on parenting self-efficacy, empirical research examining the effects of interventions with self-compassion as the dependent variable remains limited.

Narrative therapy, developed in the context of poststructuralism and postmodernism in the 1980s, is a counseling approach based on the assumption that an individual's identity is socially constructed through interactions with others [17]. Narrative therapy views individuals as experts in their own lives, helping them separate problems from themselves, recognize strengths and resources, and reconstruct preferred life stories [18]. Previous studies have reported that narrative therapy enhances identity formation, self-esteem, self-efficacy, and ego integration across diverse populations [17]. Ko and Park [19] conducted a 5-session narrative therapy program for older mothers of adult children with developmental disabilities. Through qualitative analysis, they found that participants, who initially showed self-blame and isolation, later demonstrated positive identity changes such as acceptance, self-care, and hopeful expectations for their children. These findings suggest that narrative therapy enables mothers of children with disabilities to rediscover overlooked values and strengths, thereby enhancing self-compassion, while also reframing caregiving adversities to strengthen parenting confidence and ultimately improve parenting self-efficacy [20].

Domestic intervention studies on narrative therapy are often applied eclectically with other approaches; however, limited discussion on theoretical integration or potential conflicts may weaken its philosophical foundation and unique characteristics [17]. Choi et al. [21] conducted a 12-session Narrative Therapy-Based Integrative Counseling for Group (NTBIC-G) with mothers raising children diagnosed with autism spectrum disorder, which combined narrative therapy with cognitive-behavioral and emotion-focused elements, and reported improvements in self-esteem and life satisfaction. However, this program had limitations in isolating the unique effects of narrative therapy. In contrast, Karimi et al. [22] implemented an 8-session pure narrative group counseling program for mothers raising children with intellectual and developmental disabilities and reported significant improvements in psychological capital, including self-efficacy, hope, resilience, and optimism. These findings suggest that narrative therapy can be effective as a stand-alone intervention.

Nevertheless, few experimental studies in Korea have applied pure narrative therapy to mothers of children with disabilities, and even fewer have examined self-compassion and parenting self-efficacy as dependent variables. Therefore, this study aimed to examine the effects of group counseling utilizing narrative therapy on self-compassion and parenting self-efficacy in mothers of children with disabilities.

• RQ1. Do mothers who participate in group counseling utilizing narrative therapy show significantly greater improvements in self-compassion compared to those who do not participate?

• RQ2. Do mothers who participate in group counseling utilizing narrative therapy show significantly greater improvements in parenting self-efficacy compared to those who do not participate?

This study aimed to examine the effects of group counseling utilizing narrative therapy on self-compassion and parenting self-efficacy among mothers of children with disabilities.

Hypothesis 1 states that the experimental group that received group counseling utilizing narrative therapy intervention would show a significantly greater increase in self-compassion scores than would the control group.

Hypothesis 2 states that the experimental group that received group counseling utilizing narrative therapy intervention would show a significantly greater increase in parenting self-efficacy scores than would the control group.

This study employed a quasi-experimental design using a nonequivalent control group pretest-posttest nonsynchronized design. The experimental group participated in group counseling utilizing narrative therapy, whereas the control group received no intervention.

The participants included mothers residing in City G who were raising children with disabilities. Recruitment was conducted through online and offline announcements on bulletin boards of relevant institutions, such as the Special Education Support Center, the Comprehensive Welfare Center for the Disabled, and the Parents' Association for Children with Disabilities. Eligible participants were mothers of children under 18 years of age who either held a disability registration card or had a disability diagnosis certificate and whose children were enrolled in special education institutions (such as special schools, special classes within regular schools, or specialized childcare centers for children with disabilities). Mothers who could not communicate in Korean or were currently participating in other counseling-based programs were excluded from the study.

The sample size was calculated using G*Power version 3.1.9.7 based on the effect size reported in a previous study [23] that applied group counseling utilizing narrative therapy. Assuming ⍺=.05, an effect size of .80, and power=.80, the required sample size was calculated as 21 participants per group. Considering a 10% attrition rate, 23 participants were purposively sampled in each group using convenience sampling. During the intervention period, three participants in the experimental group dropped out due to family caregiving responsibilities, whereas three participants in the control group declined to complete the posttest. Thus, the final sample consisted of 20 participants in each group (Figure 1).

Prior to participant recruitment, the researcher obtained permission from the director of the Mental Health Welfare Center in City G, where the intervention was conducted. After explaining the research purpose, recruitment materials and official cooperation letters were distributed in person and by email to relevant institutions, including special daycare centers, disability welfare centers, parent organizations, and special education institutions. The recruitment materials clearly described the study objectives, eligibility criteria, participation procedures for both groups, and the right to voluntarily join the preferred group. Recruitment announcements were also disseminated through social networking sites and offline bulletin boards. After thoroughly reviewing the study information and participation requirements, participants selected between the experimental and control groups according to their individual circumstances and preferences, and voluntarily consented to participate. Group allocation was therefore determined entirely by participants' autonomous choice, without any forced assignment between groups.

To prevent cross-contamination between groups, pretest and posttest data collection for the control group was completed before starting the intervention for the experimental group. Participants in the control group were contacted via phone, provided with study information and requested to complete the consent form and pretest accessible through Google Forms. The posttest was conducted in the same manner 6 weeks later. Before conducting the pretests and posttests for the experimental group, the researcher developed a protocol covering the program procedures, tools, and data collection. The pretest was conducted independently by research assistants trained in the study protocol. After the 12-session intervention, which was conducted over 6 weeks, posttest data were collected by the research assistant using the same questionnaire, excluding the general demographic items already obtained in the pretest.

The program was conducted from May 7 to June 14, 2024. The program was adapted to the characteristics of mothers of children with disabilities, drawing on Kim's [18] narrative therapy-based individual counseling framework and Park and Kim's [23] group counseling program for individuals with alcohol dependency. The terminology and presentation format from Park and Kim's [23] manual were revised. The program integrated the in-depth approach of individual counseling with interactive group activities, focusing on the protagonist's narrative and the reflective responses of group members. The facilitator's questions were also modified to distinguish between the socially constructed identity of 'mother of a child with a disability' and the individual's unique identity, while incorporating coping experiences in the parenting process. To assess the appropriateness of the modified program, two psychiatric nursing professors certified in narrative therapy and one psychiatric nurse specialist reviewed the content for validity. Subsequently, they found that the Content Validity Index for all eight evaluation items was 1.00, indicating high content validity.

The program structure was designed based on previous studies analyzing intervention programs for mothers of children with disabilities, which suggested that 10 to 12 sessions were most effective [13]. Therefore, the program was designed to consist of 12 sessions, each lasting 60 minutes, held 1 to 3 times per week. The group composition was based on prior studies applying group counseling programs utilizing narrative therapy [17,19,21]. Considering that the intervention was conducted individually for each protagonist, groups of 5 to 6 participants were formed to ensure sufficient time for each protagonist. Additionally, based on the preferences of the experimental group, the participants were divided into four groups: Groups A, B, C, and D, who met every Tuesday, Wednesday, Thursday, and Friday, respectively. The locations of the sessions were selected considering the participants' accessibility, with groups A and C holding their meetings at G Mental Health and Welfare Center, whereas groups B and D holding their meetings at G Mental Health Recovery Support Center.

Unlike conventional group counseling programs that center each session on a common theme for multiple participants, this program was structured to reflect the philosophical principles of narrative therapy. These principles are grounded in social constructionism, which views identity as shaped through social interaction, and emphasizes respecting clients as the experts and primary authors of their own lives. Accordingly, each participant served as a 'protagonist' for two sessions, exploring her narrative in depth and experiencing identity reconstruction and transformation. The other group members assumed the role of outsider-witness, engaging in reflective activities that reinforced the protagonist's redefined identity through shared responses and validation.

The program's overall structure was as follows (Table 1). The first week consisted of one orientation session. From the second to fifth weeks, two sessions per protagonist (covering four protagonists) were held. The sixth week consisted of two sessions for the final protagonist and one final session to close the program.

In the first week, Session 1 served as an orientation to the program, where participants were introduced to the program's purpose and encouraged to select and introduce themselves using positive nicknames to foster group rapport. The theoretical foundations of narrative therapy were introduced through handouts, followed by the group pledge on "Our Promises" and "Our Beliefs," which outlined the ground rules based on the principles of narrative therapy. The participants were guided in completing the Myers-Briggs Type Indicator (MBTI) personality inventory and a life-curve drawing exercise. Session 2 focused on "externalizing the problem" and identifying "unique outcomes." Participants named and symbolically detached themselves from dominant problems, enabling separation from their identity as a mother of a child with a disability. The session included a discussion of the problem's influence on different life domains and the discovery of moments when the participant resisted or addressed the problem. Session 3 involved "constructing alternative narratives" and "writing a new story." Participants scaffolded their strengths and redefined their identity using MBTI strengths and outsider witnessing. They reconstructed their support networks using eco-maps and shared their future goals through a definitional ceremony with a telling, retelling, and reflection sequence. Sessions 4~11 were conducted in the same format for the remaining four protagonists. Session 12 concluded the program by reading "To Myself as Precious as a Star," affirming the protagonist's identity and awarding symbolic trophies with self-care messages.

Ethical approval for this study was granted by the Institutional Review Board of K University (Approval No. KNU-2024-0097) prior to data collection, which was conducted from March 4 to June 14, 2024. Participants were thoroughly informed about the objectives, rationale, and procedures of the study prior to the baseline assessment. They were assured that involvement was entirely optional and that they could discontinue participation at any stage without disadvantage. Written consent was obtained from every participant after providing sufficient information. To safeguard privacy, all data were anonymized and handled with strict confidentiality. Participants were also notified that the information would be used exclusively for academic research and would be discarded once the study was finalized.

After the program, the experimental and control groups were given small tokens of appreciation. Inquiries regarding the study or its results were answered in good faith. Additionally, to ensure that no ethical considerations could be raised for participants in the control group, they were offered the workbook for the intervention and a condensed version of the program (two sessions) after the study's conclusion.

The program was conducted solely by the researcher, who holds certifications as a Psychiatric Mental Health Nurse Specialist (PMHNS) and a Level 1 Mental Health Nurse. The researcher has eight years of experience providing individual and group counseling as well as educational services at a Mental Health Welfare Center. In addition, the researcher regularly participates in workshops and case reflection teams organized by the Korean Narrative Counseling Association to continuously enhance professional expertise. Data collection and pre- and post-assessments were carried out by trained research assistants.

Collected data were processed and analyzed using IBM SPSS Statistics version 25. Descriptive statistics, including frequency, percentage, mean, and standard deviation, were applied to analyze the participants' general characteristics and research variables. The results of x2 and Fisher's exact tests indicated homogeneity between the experimental and control groups regarding general characteristics. The Kolmogorov-Smirnov test was conducted to assess the normality of the dependent variable before the intervention, with the results confirming that the variables followed a normal distribution. An independent samples t-test was employed to analyze differences between the groups and test the research hypotheses.

A total of 40 participants (20 each in the experimental and control groups) were included in this study. Homogeneity testing of general characteristics showed no significant differences between the groups regarding the mother's age group, marital status, education level, religion, employment status, monthly household income, number of children, satisfaction with spousal relationship, or satisfaction with the mother-child relationship. Furthermore, both groups showed no significant differences in the gender of the children with disabilities, schooling level, type of educational institution, type of disability, severity of disability, or birth order, indicating homogeneity between the two groups (Table 2).

Homogeneity testing of the research variables also indicated no significant differences in self-compassion and parenting self-efficacy between the experimental and control groups at baseline, indicating initial equivalence (Table 3).