We adopted a quasi-experimental design (two groups × two periods) within a mixed-methods approach that integrates quantitative and qualitative analyses. Quantitative data were collected by school personnel through a questionnaire comprising seven scales. Subsequently, researchers conducted semi-structured interviews to gather feedback from participants in the intervention group, given the small sample size of questionnaire responses [60]. Thematic analysis [61] was used to qualitatively assess changes in the intervention group and reveal the effectiveness of the intervention and directions for improvement.

The school-based parenting intervention in this study was implemented within the primary school setting. Researchers identified two comparable public primary schools (designated School A and School B for anonymity). At School A, 15 families were recruited and selected to participate in the intervention group. These families participated in family interviews and completed the family background and pre-test questionnaires. This data collection served three primary purposes: (1) to understand the basic information and needs of the intervention group participants; (2) to obtain necessary information for conducting propensity score matching (PSM); and (3) to establish baseline measurements for the intervention group. PSM was employed to create balanced groups with similar distributions of observed covariates [62].

The control group was selected using PSM from a larger pool of second-grade families (n = 223) at School B. These families also completed the family background questionnaire used for the PSM procedure. Successfully matched families did not participate in the intervention but completed both pre- and post-test questionnaires concurrently with the intervention group. Cross-school matching not only improves comparability between the intervention and control groups and eliminates confounding factors in the experimental results [63], but also prevents spillover effects from parent communication within schools [64,65]. This approach ensures that the assessment of the intervention effect has high internal validity.

This study was approved by the Shaanxi Education Science Research Institute (Approval No.: SGH20Y1016; Date: 15 January 2021), which conducted a full ethics review as part of the project authorization process. This procedure is equivalent to that of the Institutional Review Board in the United States. All procedures exclusively involved adult parents; no minors participated directly in interventions or data collection. Written informed consent was obtained from each participant prior to enrollment, documenting the study’s purposes, procedures, and rights. Regional education authorities supervised implementation in compliance with data confidentiality regulations.

To implement family-school co-education, a county education bureau and the author jointly developed this school-based parenting program to enhance parenting quality. Two public primary schools (School A and School B) were selected based on comparable demographics, including location, student enrollment, second-grade cohort size, teacher-student ratio, and teacher qualifications (Appendix A). Intervention group participants were recruited from families (n = 352) of second-grade students in School A. Inclusion criteria required: (a) Children aged 7–9 years; (b) No diagnosed behavioral/psychological problems; (c) Primary caregiver parents; (d) Parents without diagnosed psychological/behavioral illness; (e) ≥1 parent able to attend all sessions; (f) Parental literacy for curriculum tasks. Fifteen families were selected from a pool of 18 recruits to form the intervention group. The study experienced limited attrition, with two families withdrawing due to time constraints. The final sample (n = 13 children, aged 7–8 years) aligned with the recommended group sizes for parenting interventions [66] (Table A1).

The control group was drawn from second-grade families (n = 223) in School B. We excluded children with health problems, non-parental households, or divorced/separated parents. Propensity score matching (PSM) created 1:1 pair-matched families from Schools B and A.

Propensity scores were calculated using eight family background variables [67] (n = 212), including the child’s age, gender, family structure, number of children, parental employment, household income, and parental education (both the father’s and mother’s). Using logistic regression and nearest-neighbor matching (caliper = 0.25 SD [68]), we obtained a balanced sample of 30 families (School A = 15, School B = 15) prior to attrition. The matching procedure was successful, with no significant differences in baseline characteristics between two groups (see Appendix B) (Table A2).

Post-matching, two intervention-group families withdrew before the baseline assessment due to scheduling conflicts (confirmed through follow-up contact). Baseline comparisons revealed no significant differences among the remaining groups, confirming that attrition did not affect group comparability or introduce systematic bias. Sample screening is detailed in Fig. 1, and group characteristics are presented in Table 1.

The intervention outcomes were evaluated based on quantitative data collected using a standardized questionnaire that contained seven scales measuring CEBP, parental adjustment, parenting confidence, parent-child relationships, parenting style, parental relationship, and parenting conflict. All the scales were administered in their Chinese versions, an English PhD expert performed meticulous translation checks to ensure the participants’ full understanding of the questionnaire items. Before distributing the questionnaire, feedback was solicited from a pilot sample of 20 participants, leading to further refinements in the language and wording of the scales.

The school-based parenting intervention aimed to help parents adopt authoritative parenting concepts and develop new interaction strategies to improve parent-child relationships and guide children’s behavior. Our curriculum framework integrated attachment theory and social learning theory through two dimensions (positive interaction and firm guidance) and four components (relationship bonding, emotional support, behavior training, and behavior management). The positive interaction dimension emphasized emotional bonding while teaching practical skills for emotional expression and parent-child engagement, countering the emotional restraint typical in traditional Confucian parenting. The firm guidance dimension focused on identifying the underlying logic and needs of children’s behaviors, offering non-coercive practical skills to build positive habits and manage challenging behaviors. This explicitly targeted parent-child conflicts during socialization, particularly academic conflicts frequently experienced by Chinese parents.

The curriculum content encompassed concept introduction (such as positive and negative parenting styles), skill training (such as descriptive praise, non-violent communication, and natural and logical consequences), and case-based practice (e.g., tutoring homework, family dinner, and addressing mobile phone addiction). Moreover, we used participant co-design approaches and post-session assignments to capture practical feedback on parenting skills in real-life situations. To ensure effective delivery, we monitored the intervention process through attendance rates, in-class participation rates (measured by active questioning), and post-session feedback (as indicated by questionnaire completion) (Appendix C). The intervention was designed and implemented by a team of professional social workers, including two university professors (the first and third authors) and one PhD candidate (the second author). All team members received comprehensive training in the fundamental theories of child and family studies, as well as social work group facilitation skills (Table A3).

The curriculum consisted of seven sessions involving four modules (three positive parenting courses, one parents co-design course, two parenting practice courses, and one final course) (Table 2). Participants visited Y school for seven group sessions once per week for 1–2 h (the last four sessions moved online due to COVID-19 stay-home orders). Following the intervention, 10 participants participated in semi-structured interviews to discuss their experiences with applying skills. One-to-one semi-structured interviews were conducted via video conference by the first author, with the second author documenting the interviews. Each 60-minute session required prior informed consent and permission to audio-record. Interview topics explored intervention objectives, parenting skill application, family changes, and program experiences (Appendix D). All procedures concluded in February 2022 (Table A4).

The quantitative data were analyzed using a t-test, chi-square test, and ANCOVA. The samples were found to be normally distributed and independent. Independent sample t-tests and chi-square tests were used to test the balance of the variables between the intervention and control groups. Paired-sample t-tests and ANCOVAs were used to test the within- and between-group differences between the intervention and control groups. Effect sizes (Cohen’s d) were interpreted as small (0.2), medium (0.5), and large (0.8), per convention. In addition, we conducted the non-parametric Mann–Whitney–Wilcoxon test to verify the robustness of the results. The quantitative data were analyzed using SPSS 26 (IBM Corp., Armonk, NY, USA).

The qualitative data were examined using thematic analysis, which is “a method for identifying, analyzing, and reporting themes in data” and is a valuable and reliable qualitative research method [61]. Two principal investigators (the first and second authors) independently conducted line-by-line coding of transcribed interviews, generating 126 initial codes through semantic coding of explicit behavioral descriptions (e.g., “praised after packing up the schoolbag”) and latent coding of relational dynamics (e.g., “child shared his secrets”). Through three consensus discussions, the codes were grouped into 21 subthemes, with an 86% agreement rate. Discrepancies were resolved through a collective re-examination of the raw data, with the third researcher (the third author) verifying inconsistent themes. Through thematic integration, five themes were identified to reflect the core concepts across the interviews. The three researchers reached a consensus on the final themes, ensuring the credibility of the thematic analysis process.

Five themes related to the effectiveness of the intervention were identified from the 10 semi-structured interviews: (1) perceptions of parenting behaviors, (2) awareness of children’s good behavior and misbehavior, (3) practice of positive parenting skills, (4) growth of parents and children, and (5) suggestions for programs and interventions. These themes are elaborated upon in the following sections, supported by direct quotations from the interview data.

This pilot study represents one of the limited school-based parenting intervention trials conducted in mainland China. Such interventions remain relatively scarce compared to Western counterparts or clinic/community-based programs in China [44–48]. The quantitative results indicated that the positive parenting intervention significantly improved CEBP, parenting confidence, parent-child relationships, and parental adjustment. However, no significant effects emerged for parenting style, parental relationships, or parenting conflict. Qualitative findings further revealed improvements in CEBP, parental confidence, parent-child relationships, parenting stress, and parental perceptions and behaviors.

These findings align with a recent meta-analysis demonstrating that positive parenting programs prevent childhood socioemotional-behavioral problems by enhancing parental confidence, adjustment capacity, and parent-child relationships [84]. Neither this study nor the meta-analysis detected significant intervention effects on parental relationships or reductions in parenting conflict. Although consistent with prior research, the small sample size may constrain generalizability. To bolster reliability and mitigate within-school spillover effects, we employed a quasi-experimental PSM design across two schools [63].

The intervention and post-test coincided with COVID-19 lockdowns, imposing sustained familial stressors. Final sessions transitioned online due to lockdown restrictions. Protocol adaptations (e.g., real-time chat, interactive modules) ensured the fidelity of the online intervention. This shift maintained attendance and enhanced engagement, with fathers notably utilizing chat functions to voice concerns (see Appendix C), demonstrating online delivery’s potential to enhance accessibility. One systematic review further confirms that online parent interventions during the COVID-19 pandemic achieve comparable efficacy to in-person delivery while maintaining high acceptability [85].

Empirical evidence from China and Singapore indicates that social distancing measures significantly increased parenting stress, disrupted parent-child interactions, and exacerbated CEBP, particularly anxiety, irritability, and hyperactivity [86–88]. Intervention group participants further reported pandemic-related distress in their post-class feedback, such as disrupted daily routines, remote-learning challenges, and adverse parenting practices. The control group showed significant negative changes in CEBP, parenting confidence, and parent-child relationships. Conversely, quantitative and qualitative evidence indicated meaningful improvements in parental stress reduction, parent-child relationships, and parenting confidence within the intervention group, despite non-significant changes in CEBP. This finding aligns with evidence that universal prevention programs primarily reduce risk factors for disruptive behavior (e.g., harsh parenting) while enhancing developmental outcomes [39].

Qualitative data provided critical insights for program refinement. Although quantitative measures did not capture changes in parenting style, qualitative findings report documented behavioral improvements. This discrepancy may stem from the intervention’s brief duration, which is insufficient to alter established practices. Owing to Confucian cultural influences on parenting practices, short-term interventions (<6 months) often fail to facilitate new skill implementation [42]. Universal interventions additionally face floor effects, where no significant problems at baseline limit measurable reduction potential [39].

Measurement limitations may also explain null findings for parenting style. The Parenting Style scale primarily assesses coercive and inadequate behaviors rather than comprehensive practices. In collectivist cultures like China, self-reports are vulnerable to social desirability bias [89], and school authority pressures may have inflated parenting self-assessments, undermining validity. Thus, this instrument lacks sensitivity for detecting short-term changes in community samples.

Future studies should employ measurement tools validated for ordinary Chinese, such as the Parenting and Family Adjustment Scale [44,78]. Culturally grounded measure development remains imperative, as Western constructs may incompletely capture China’s unique parenting dimensions [27]. Finally, multi-informant data (e.g., from both parents and teachers) could capture comprehensive perspectives. Consequently, extending the intervention duration while adopting sensitive assessment tools would better reveal the effects of parenting style.

Parental co-involvement is crucial for enhancing relational dynamics and mitigating parenting conflict. Although quantitative results were non-significant, deteriorating parental relationships in controls and qualitative reports underscore the necessity of dyadic engagement in conflict resolution. Despite the active recruitment of both parents, co-attendance proved challenging. The parents in the intervention group (usually mothers) reported limited spousal and grandparental support when implementing parenting skills, which exacerbated intergenerational conflicts and marital strain.

Family system theory [90] posits that families function through interdependent subsystems, where spousal cooperation underpins systemic stability, and alterations in one subsystem cascade through others. Consequently, interventions targeting individual parents struggle to sustain behavioral change [91]. Extensive evidence confirms the critical role of paternal involvement in establishing co-parenting relationships, influencing caregiver mental health, parental and parent-child relationship quality, intergenerational transmission patterns, and balancing extra-familial stressors [92,93]. Despite their developmental significance, Chinese fathers’ traditional provider role (vs. caregiving expectations) [94] and rigid paternal identity norms systematically impede participation [93–95]. Global evidence confirms that paternal warmth, equity, and non-violence enhance family well-being [96]. Additionally, frequent grandparental involvement—substantially exceeding Western levels—and resultant intergenerational conflict intensify implementation barriers [56,97].

Thus, promoting collaborative parenting and strengthening father involvement are essential for addressing these challenges in China. Evidence confirms that grandparent-inclusive interventions enhance caregiver mental health and reduce child behavior problems via structured relational components [98,99]. Key strategies to promote father engagement include (1) adopting strength-based paternal frameworks, (2) customizing content and delivery for fathers, and (3) training practitioners in father-inclusive methods [100]. Researchers should further explore consensus-building mechanisms adapted to China’s sociocultural context.

Several limitations warrant acknowledgment. First, the small county-based sample limits generalizability to broader Chinese populations. Future studies should verify the effects using larger, more diverse samples. Second, the absence of an RCT design limits causal inference, although the PSM-enhanced quasi-experimental design partially mitigates this limitation. Future work should prioritize RCTs. Finally, the lack of follow-up assessments prevents longitudinal evaluation. While this study focused on immediate validation and program optimization, future research should extend observation periods to capture the evolution of parenting style and the long-term impacts on the family, given the gradual nature of behavioral change. The objective of this study was to explore the effectiveness and feasibility of positive parenting intervention within a specific context. Future research efforts should further investigate this area to establish a strong evidentiary foundation.

These findings carry significant theoretical and practical implications. Results indicate that authoritative parenting frameworks are adaptable to Chinese families; however, participant heterogeneity in education and ability may challenge the efficacy of group interventions. Complex strategies (e.g., natural/logical consequences) demand precise situational sensitivity, suggesting future programs should incorporate: (1) extended duration, (2) tiered skill-building modules, (3) multimedia instructional supports (e.g., detailed manuals, instructional videos), (4) integrated case services, and (5) differentiated intervention groups based on participant characteristics. Moreover, supplementary modules addressing parental relationships and intergenerational conflict resolution may facilitate paternal participation and address distinctive parenting risks in China, particularly intergenerational caregiving challenges. The study further confirms the strategic value of school-based interventions with co-design components, where schools provide critical infrastructure for parenting programs [46] by enabling curriculum standardization tailored to primary school parents, ensuring robust participation through institutional channels, and sustaining support networks via school-linked parent communities. Significantly, the integration of group and case models proves effective for service delivery. Group formats create supportive environments that enable the non-judgmental sharing of concerns, peer interactions foster perspective transformation through the exchange of experiences, and case services provide individualized solutions for family-specific needs [2]. Practitioner flexibility remains essential, where collaboration with parents through formal interviews, course interactions, and feedback mechanisms enhances program responsiveness to local contexts while preserving core elements [10,101]. Consequently, future research should prioritize school-based settings, diversify intervention methods, and expand participatory co-design to identify optimal implementation pathways for effective interventions.