This study used a mixed methods, pretest-posttest feasibility design with a prospective cohort of participants enrolled in the SF Program. Phase 2 ran a follow-up data collection at 4 time points (before the intervention and at 2, 6, and 12 weeks after the end of the intervention). The pretest was completed prior to the first week, and the posttest was completed within 12 weeks of the final programming date for estimating the long-term outcomes. Additional qualitative interviews were conducted between August 2025 and September 2025, approximately 12 weeks following the last session. Figure 1 illustrates the study timeline.

This program was conducted between June 2025 and July 2025 in the Chiang Mai, Chiang Rai, and Lamphun provinces in northern Thailand, encompassing 8 study locations in total. Training for the Trainers of the SF Program was a 3-day workshop for trainers who would conduct SF in their local geographic areas. The workshop was led by master trainers certified by UNODC. The SF Program consisted of 3 sessions held on 3 consecutive weekends engaging caregivers and children. The first session was exclusively for caregivers, whereas children participated in parallel and joint activities in the second and third sessions, respectively. The topics for caregivers were (1) understanding strengths and stressors, (2) using love and limits, and (3) teaching children what is right. The topics for children were (1) learning about stress and (2) following rules and appreciating parents. The topics for joint sessions were (1) learning about each other and (2) supporting values and dreams. To ensure implementation fidelity, UNODC-certified master trainers attended and provided feedback for the Thai trainers at community sites during the first session of the SF Program.

The target population for this study consisted of the families (specifically, the caregivers and their children) who participated in the SF Program phase 2. We used a convenience sampling strategy recruiting 68 volunteer families (caregiver-child dyads) from 8 study locations across the 3 provinces. Families with children aged 8 to 15 years were included, aligning with the program’s target demographic.

This study was conducted following review and approval from the Institutional Ethics Committee of the Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand (ethics approval number 143/2025). All participants provided written informed consent prior to study commencement in accordance with the 1964 Declaration of Helsinki.

A total of 68 families attended the SF Program across the Chiang Mai, Lamphun, and Chiang Rai provinces in 2025. Most caregivers (59/68, 86.8%) were female, with a mean age of 47.4 (SD 13.8) years. In total, 51.5% (35/68) of the children were male, and their mean age was 10.8 (SD 1.6) years.

Table 1 shows the scoring and changes in all assessment tools. Regarding the PAFAS, the results showed a significant improvement in coercive parenting (β=–0.10; P=.003) as the decrease in scores reflected a reduction in maladaptive behaviors. However, the analysis indicates a significant increase in scores for positive encouragement (β=0.05; P=.009) and parent-child relationship (β=0.28; P<.001); based on the inverted scoring key, these suggest a decline in positive parenting skills and relationship quality during the study period. In terms of child difficulties (evaluated using the SDQ), parents reported significant improvements, evidenced by decreases in conduct problems (β=−0.04; P=.002) and peer relationship problems (β=−0.04; P=.048). However, social relationships (a strength-based domain) showed a decrease in scores (β=−0.05), suggesting a reduction in strengths, although this was not statistically significant (P=.09). Similarly, children self-reported significant improvements in emotional (β=−0.05; P=.01) and conduct (β=−0.04; P=.02) behaviors. No substantial changes were observed in child resilience (CYRM-R-PMK) from either the parent or child perspectives.

The qualitative study included a total of 19 families, or caregiver-child dyads. The participating caregivers were predominantly female (17/19, 89.5%), comprising 82.4% (14/17) mothers, 11.8% (2/17) grandmothers, and 5.9% (1/17) aunts, whereas the 10.5% (2/19) male participants included 1 father and 1 foundation staff member (caregiver for a child in the foundation). The ages of these caregivers ranged from 28 to 59 years, representing early working age to middle adulthood. Regarding the children, 52.6% (10/19) were female, and 47.4% (9/19) were male, with ages ranging from 7 to 14 years. This age group corresponded primarily to the late primary and early secondary school levels, critical developmental stages (early adolescence and middle childhood) for behavior and responsibility. From the qualitative analysis, 5 themes emerged (Textbox 1).

The main motivation for enrolling in the program was the desire to improve parenting skills. Caregivers indicated that they wanted to learn the correct ways to teach and manage their children’s behavior and understand their children better. Some caregivers also sought to resolve relationship issues at home, motivated by a desire to reduce family stress or strengthen the bond between mother and children. Other motivations for caregivers and children to participate in the program were varied, including persuasion from community and school leaders; most participants were referred by village heads, schoolteachers, or religious leaders (churches and temples). Finally, some children expressed interest in enrolling so that they could participate in the children’s activities, either as a way for the latter to enjoy time with their parents or as a response to an invitation from friends:

Family members reported acquiring new skills and practices, including stress management techniques such as breathing exercises and somatic coping techniques (ie, tapping various points on the body) to help control anger and reduce stress. Caregivers also reported learning positive communication skills, including encouragement to use reason instead of emotions, speaking in a softer tone of voice, and actively listening to children’s needs. Caregivers shared that they learned about their children’s stress levels through activities such as the “stress ladder” or “thermometer,” gaining insights into children’s perspectives that adults might not have previously known. They also reported learning how to show affection and establish discipline appropriately without using violence, including techniques for hugging, praising, and setting rules in the home:

Caregivers shared insights that they became more mindful and patient and reduced scolding, yelling, and use of violence (hitting children). Instead, they started reasoning more. Children became more responsible, enacting behaviors such as helping with household chores (washing dishes, doing laundry, and ironing), studying diligently, and doing homework independently without being told. Additionally, some children showed a decrease in aggressive behavior, such as stopping or reducing bullying at school. At the family level, it was found that the atmosphere at home improved; members became closer, talked more openly, and showed more affection toward each other such as hugging or kissing cheeks. Positive changes extended to other family members even though the spouses or other parents did not attend the training. When the parent who attended the program changed their behavior, the nonattending parent was found to drink less alcohol or act less violently and be more helpful with household chores. The younger siblings who did not attend the program often followed the positive behavior of the older sibling who did participate:

Caregivers commented that the SF Program was a creative and fun activity and a family learning activity through shared experiences. Hearing problems from other families helped parents see their own strengths and weaknesses, not feel alone in solving problems, and know that they had clear and effective tools to improve the functioning of their family:

To improve the experience of being enrolled in the SF Program and the family outcomes, caregivers stated that they would like to see activities organized regularly or continuously (eg, once a year) to prevent knowledge from fading. Participating caregivers suggested that the target group should be expanded to include more family members, such as other relatives, guardians, or siblings. Regarding the venue and time, participants shared that the activities were too short and some lacked depth. Some families suggested holding the SF Program sessions on Saturdays (as opposed to Sundays) to avoid religious obligations, and they also wanted a larger venue. They would like to see more content and training on etiquette for children, the language in the curriculum simplified, and families with specific problems selected as the target group to achieve the most noticeable changes:

This study represents the evaluation of the UNODC SF Program in Thailand. Findings suggest that while the program was highly acceptable and qualitatively perceived as beneficial, the quantitative outcomes present a complex picture of family dynamics during the transition phase of behavior change. The pattern of findings observed in this study is consistent with family systems and developmental prevention theories, which emphasize the reciprocal and dynamic nature of caregiver-child interactions. Improvements in coercive parenting alongside reductions in child conduct problems suggest that even brief changes in caregiver behavior can rapidly influence child outcomes through feedback processes within the family unit. Family skills interventions that focus on caregiver stress regulation, emotional awareness, and nonviolent discipline have been shown to interrupt maladaptive interaction cycles and promote healthier relational patterns, particularly in contexts of chronic socioeconomic stress. In this way, the SF Program operates by modifying discrete behaviors and also by strengthening relational processes that support child development and family resilience within broader ecological systems [4,18,19]. Regarding parenting and family adjustment (PAFAS), the results showed a significant improvement in coercive parenting. However, a decline in positive parenting skills and relationship quality during the study period was observed. In terms of child difficulties (SDQ), parents reported significant improvements. However, social relationships showed a non–statistically significant reduction in strength. Similarly, children self-reported significant improvements in emotional and conduct behaviors. No significant changes were observed in child resilience (CYRM-R-PMK) from either the parent or child perspective.

These findings are consistent with the program’s logic model, which emphasizes stress management and nonviolent discipline. The qualitative data strongly corroborate this as caregivers reported becoming more mindful, reducing yelling, and using reasoning instead of violence in interactions with their children. Similarly, children self-reported significant improvements in emotional and conduct behaviors, suggesting that even a short 3-week intervention can interrupt the cycle of negative interaction within the family. Additionally, our study on the early implementation of the SF Program in Thailand, which assessed feedback from the trainers of SF, also illustrated this consistent outcome of improving family function [8]. Previous literature illustrates that positive parent-child relationships, especially early and middle childhood, are associated with a favorable impact on children’s behavior, avoid early-onset conduct issues, and act as a buffer against negative childhood experiences, which reduces toxic stress and improves health [20,21].

An intriguing finding in our study was the discordance between the quantitative scores for parent-child relationship and positive encouragement and the qualitative narratives. While the quantitative analysis indicated a significant increase in scores, interpreted as a decline in skills based on the instrument’s inverted scoring, the qualitative themes painted a contrasting picture. In the interviews, families reported improved atmospheres, increased affection (eg, hugging), and better communication. The discordance may be attributed to response shift bias [22]. Participants likely recalibrated their internal standards of positive parenting after the intervention, leading to stricter self-evaluation at the posttest compared to the pretest time point. This may also be viewed through the stages of competence development [23]. Participants moving from unconscious incompetence to conscious incompetence tend to judge their past and current behaviors more critically, resulting in lower postintervention scores despite actual skill acquisition [23]. The significant reduction in child conduct problems (SDQ) contrasted with the nonsignificant changes in resilience scores (CYRM-R-PMK). This suggests a distinction between acute behavioral symptoms, which respond rapidly to changed parenting strategies, and resilience, which is a complex, enduring trait requiring a longer duration or sustained intervention to manifest significant change [24]. Furthermore, qualitative insights revealed that caregivers reported uncertainty about their children’s expectations, suggesting a period of relational recalibration and heightened sensitivity rather than deterioration. These findings underscore the necessity of the mixed methods approach adopted in this study as it captured the nuanced developmental processes that quantitative tools alone might overlook [25].

Rather than reflecting program failure, the observed response shift in self-reported parenting measures may represent a meaningful indicator of learning and engagement. Response shift occurs when participants recalibrate their internal standards or conceptual understanding following exposure to an intervention, resulting in more critical self-assessment at the posttest time point despite genuine behavioral improvement [22]. In parenting interventions, increased awareness of positive parenting norms may lead caregivers to judge their behaviors more stringently over time. This process aligns with stages of competence development, in which individuals move from unconscious to conscious incompetence before achieving mastery [23]. Qualitative reports of increased patience, reflective communication, and emotional regulation suggest that caregivers were actively internalizing program content, underscoring the value of mixed methods designs for capturing developmental processes that may not be fully reflected in standardized self-report measures alone [25].

The nonsignificant resilience findings also contrast with rapid gains reported in conflict-affected settings that implemented and evaluated SF, such as Afghanistan and Iran [26,27]. This divergence likely reflects baseline severity; populations with high trauma exposure, such as those in prior SF implementations, often have lower baselines and greater room for improvement. In the nonconflict Thai cohort, where stressors were socioeconomic rather than acute, resilience building appears to be a gradual process requiring a longer intervention duration than 3 weeks to produce statistically detectable change.

Thai communities accepted SF in the 2024 pilot, with trainers reporting that the curriculum improved family relationships and mutual understanding. The pilot’s qualitative comments revealed language obstacles. Facilitators also struggled to handle intensely tough, highly emotional familial situations that spontaneously developed during group sessions. The second phase allowed for the incorporation of brief interventions and strategies for handling difficult situations relevant to the local context, as suggested from our previous phase [8]. Thus, program materials were culturally localized before phase 2. The translated texts were heavily simplified into colloquial Thai; visual assets were replaced with culturally congruent imagery (eg, Asian faces and contextually appropriate domestic scenarios); and, most importantly, the Training for the Trainers curriculum was expanded to teach trainers how to offer targeted advice and de-escalate difficult participant interactions during live facilitation. From families’ perspective, this study confirms that the SF Program is feasible and culturally adaptable for Thailand. Qualitative feedback highlighted that the program was viewed as creative and fun, and the shared experience with other families helped reduce isolation. The motivation to join varied from community persuasion to a genuine desire to resolve family conflicts, underscoring the importance of community leaders (eg, village heads and teachers) in the recruitment process for such public health interventions. The research shows that implementing SF was feasible and had a good impact on parenting skills and the mental health of families. Because the program includes sessions for both caregivers and children independently and together, SF can enhance positive characteristics and understanding of each other and help families develop new skills in coordination [27]. Families can benefit from a wide range of positive parenting activities, such as the development of good relationships, improved communication, and stress management [28].

In Myanmar, the UNODC executed a vital implementation of the SF Program in Kachin State. This open, multisite pilot feasibility and acceptability trial enrolled 100 families using a universal approach within the camps. The program demonstrated remarkable efficacy. There was an improvement across the SDQ regarding child mental health, as well as distinct enhancements in parenting practices and parent and family adjustment skills. Crucially, the data indicated that improvements were most pronounced among caregivers and children who presented with the greatest psychological and behavioral challenges at the baseline assessment, proving the intervention’s high utility in acute humanitarian crises. The success of this trial directly advocated for the programmatic prioritization of family skills [29].

We observed no significant changes in child resilience (CYRM-R-PMK). Resilience is a complex, multidimensional construct that likely requires a longer duration to manifest than the immediate behavior changes observed in conduct or parenting styles. To increase resilience, the intervention needs to be continuous and comprehensive [30,31]. A 3-week intervention may plant the seeds for resilience, but measurable change may require longitudinal follow-up or sustained booster sessions, as suggested by participants who requested continuous activities to prevent knowledge from fading. Additionally, a recurring theme was the need for paternal involvement. Most participating caregivers were female (59/68, 86.8%), and participants explicitly expressed a wish for husbands or fathers to attend. The predominance of female caregivers in our study reflects traditional gender roles in rural Thai society, where child-rearing is often viewed as the maternal domain [32]. However, qualitative feedback explicitly calling for paternal involvement suggests a shifting awareness within families. Future iterations of the program should specifically target fathers through gender-sensitive recruitment strategies to maximize the systemic impact on family dynamics.

Evidence from parenting intervention research indicates that father-inclusive approaches are associated with greater and more sustained improvements in child behavior and family functioning, particularly when interventions jointly address discipline practices and emotional communication [33]. Future iterations of the SF Program may benefit from gender-sensitive recruitment strategies that actively engage fathers and other caregivers, thereby enhancing the systemic and equitable impact of family skills interventions.

Several limitations should be noted. First, the sample size was relatively small (n=68 for quantitative assessment and n=19 for qualitative assessment) and restricted to specific provinces, limiting generalizability. Second, the absence of a control group prevents us from attributing changes solely to the SF intervention. Third, volunteer families may already possess higher motivation to improve family dynamics. The potential for response shift bias in self-reported measures suggests that future studies should include observational metrics or third-party reports to validate parenting changes. Fourth, there was a potential for recall bias in the interview session. However, we aimed to assess family behavior after the training session. Finally, the short follow-up period (12 weeks) may not capture the long-term sustainability of the skills acquired.

The SF Program in Thailand demonstrates strong feasibility and promising effectiveness in reducing coercive parenting, child conduct problems, and child emotional problems. While quantitative measures of relationship quality yielded mixed results, qualitative evidence strongly supports the program’s value in fostering family bonding and emotional regulation. Scaling up this intervention with an emphasis on engaging fathers and providing long-term support could significantly benefit vulnerable families in the region.