Associations between family functioning, psychological resilience, and emotional competence among primary and secondary school students in Chengdu, Sichuan Province: an exploratory study using structural equation modeling

Abstract

Background

In recent years, with rapid societal changes and increasing educational pressures, the mental health of primary and secondary school students has garnered significant attention. Psychological resilience, as a core capacity for coping with adversity, and emotional competence, as a foundation for emotional regulation in social adaptation, are crucial for student development, with family functioning being a primary environmental factor closely associated with them. Research suggests that healthy family functioning may be associated with higher psychological resilience and emotional competence, while family dysfunction may be linked to increased psychological distress. However, the interconnected mechanisms among family functioning, psychological resilience, and emotional competence, as well as the roles of factors such as gender, urban-rural differences, and grade level, still require further exploration.

Objectives

To explore the associations between family functioning, psychological resilience, and emotional competence among primary and secondary school students in Chengdu, Sichuan Province, and their underlying mechanisms. The study aims to provide a scientific basis for educators and parents to develop targeted mental health interventions.

Design

Multicenter cross-sectional study.

Methods

A cluster sampling method was employed to survey 7,937 students from grades 1 to 9 across five schools in Chengdu. Data were collected using the Chinese Family Assessment Instrument (C-FAI, assessing family mutual support, communication, and conflict harmony), the Resilience Subscale and Emotional Competence Subscale of the Chinese Positive Youth Development Scale (CPYDS, measuring adaptation and recovery under stress, and the ability to perceive, understand, and manage emotions, respectively). Data were double-entered and verified using Epidata 3.1. SPSS 26.0 was used for descriptive statistics, correlation analysis, and difference tests (independent samples t-test, Welch t-test, one-way ANOVA, or Welch ANOVA based on data distribution, with Games-Howell post-hoc tests). Partial correlation analysis controlled for gender, urban/rural residence, and grade. Structural equation modeling was conducted using AMOS 26.0 to analyze the associations and mediating effects among family functioning, psychological resilience, and emotional competence, and to evaluate model fit. Harman’ s single-factor test was applied to detect common method bias. The significance level was set at α=0.05.

Results

Family functioning showed significant differences across gender, urban/rural location, and grade level (P<0.05): male students (1.97±0.74), rural students (1.97±0.73), and students in grades 7-9 reported more severe family dysfunction. Psychological resilience was significantly positively correlated with emotional competence (r=0.646,P<0.001), and both were negatively correlated with family dysfunction (r=−0.394 and r=−0.376, respectively, P<0.001). The structural equation model demonstrated a good fit (CMIN/DF=6.988,RMSEA=0.027). Path analysis from the model indicated that psychological resilience may be indirectly associated with emotional competence through family functioning. The mediating effect of this path was 0.089, accounting for 9.2% of the total effect (95% CI: 0.667-0.726, P<0.001).

Conclusions

Family dysfunction is significantly negatively correlated with the psychological resilience and emotional competence of primary and secondary school students, with psychological resilience indirectly associated with emotional competence through family functioning. Boys, rural students, and those in grades 7–9 exhibit more severe family dysfunction, warranting focused attention. It is recommended to implement interventions such as “Parent-Child Co-Creation Day,” communication training, and “Father’ s Role Workshop” to optimize family functioning, thereby supporting the mental health of primary and secondary school students.

Introduction

With the rapid development of modern society, increasing competition, complex family structures, and rising parental expectations for their children, mental health issues among adolescents have become a focal point of societal concern [1]. According to the World Health Organization, approximately one in seven (14%) adolescents aged 10 to 19 worldwide suffer from mental disorders, accounting for 15% of the global disease burden in this age group. If these issues are not addressed promptly, they may persist into adulthood, impairing both physical and mental health and limiting opportunities for a fulfilling life [2]. In China, the mental health problems of primary and secondary school students are also a significant concern. Studies show that about 13% of primary school students, 15% of middle school students, and 19% of high school students face psychological and behavioral issues, with mental health problems becoming more prevalent at younger ages. The severity and ubiquity of these issues require urgent attention [3, 4]. During their growth, primary and secondary school students face multiple psychological challenges, including academic pressure, family conflicts, difficulties in social adaptation, self-identity confusion, and insufficient emotional regulation skills [1, 5, 6]. These factors are intertwined and closely associated with students’ psychological development and overall well-being.

Against this backdrop, finding effective ways to alleviate the psychological stress of primary and secondary school students has become an urgent issue. Existing studies suggest a close relationship between improving emotional competence and reducing psychological stress. Emotional competence refers to the comprehensive ability to perceive, understand, and manage emotions, including identifying one’s own and others’ emotions, forming emotional connections (such as empathy and sympathy), and adapting to and responding to emotional challenges in different situations [7]. Further research indicates that through emotional regulation strategies such as cognitive reappraisal, students can reassess the meaning of stressful situations, thereby reducing the intensity of negative emotions, enhancing their ability to cope with stress, and promoting both mental health and academic performance [8]. Gross and John’s emotion regulation theory highlights that cognitive reappraisal can effectively reduce stress [9], while Aune et al.‘s study indicates that emotion regulation ability mediates the relationship between stress and depressive symptoms, suggesting that emotional competence may be closely associated with reducing psychological stress among primary and secondary school students [10]. Particularly for students in the critical stage of emotional development and self-regulation formation, the cultivation of emotional competence is of utmost importance.

However, the development of emotional competence does not occur naturally but is associated with multiple factors, among which psychological resilience and family functioning stand out as core variables, playing critical roles in helping individuals cope with stress and build support systems. Psychological resilience refers to the ability of adolescents to “bounce back” from adversity and achieve healthy adaptation in the face of life stressors and developmental changes [11]. Research indicates that psychological resilience levels vary due to individual differences and are associated with factors such as student personality, psychological traits, genetic factors, cultural background, and social support [12]. Students with higher psychological resilience are often associated with stronger abilities to cope with academic stress, better interpersonal performance, improved family and peer relationships, and greater maturity in handling social conflicts [13, 14]. On the other hand, family functioning reflects the overall quality of the family environment and the interrelationships among its members [15]. Parenting styles, the closeness of parent-child relationships, the emotional atmosphere of the family, and parents’ own emotion regulation abilities are all closely associated with children’s mental health; healthy family functioning is related to the cultivation of emotional competence and the expression of psychological resilience [1, 16–19].

Based on the above research background, psychological resilience, as a core resource for primary and secondary school students in coping with stress, may be associated with the development of emotional competence through family functioning. Existing evidence suggests that higher levels of psychological resilience are related to enhanced emotion regulation abilities, which in turn are associated with alleviating the negative emotions caused by stressful life events [10]. Bronfenbrenner’s ecological systems theory posits that adolescents actively shape their family microsystem through individual traits like psychological resilience, such as by improving family functioning (e.g., communication, emotional support) through effective communication or emotional regulation, thereby promoting emotional competence [20]. Ungar’s social ecological perspective further emphasizes that resilience is a quality of the social ecology, with resilient adolescents optimizing family functioning by leveraging family resources (e.g., supportive interactions), thus enhancing emotional competence [21]. Bowen’s family systems theory also supports that resilient adolescents strengthen family cohesion by regulating family interaction patterns, providing support for emotional competence development [22]. Meanwhile, family functioning, as an external support system, may serve as a mediator between psychological resilience and emotional competence [23, 24]. However, how psychological resilience is associated with emotional competence through family functioning, and the specific mechanisms of this pathway, require further empirical research to be clarified. Given that this study is exploratory and that there is limited literature on the associations between family functioning and adolescent psychological outcomes in the Chinese context, a cross-sectional design with a large sample (N = 7,937) was adopted to establish preliminary associations, laying the foundation for future longitudinal studies. Therefore, this study, utilizing structural equation modeling, proposes the following hypotheses: (1) Psychological resilience is positively correlated with emotional competence (H1); (2) Psychological resilience is positively correlated with family functioning (H2); (3) Family functioning is positively correlated with emotional competence (H3); (4) Family functioning mediates the relationship between psychological resilience and emotional competence, with psychological resilience being positively associated with emotional competence through family functioning (H4). The hypothesized model is presented in Fig. 1.

Fig. 1.

hypothesized model

This study focuses on primary and secondary school students in Chengdu, Sichuan Province, employing a cluster sampling method to investigate the role of family functioning in the development of psychological resilience and emotional competence, as well as its underlying mechanisms. The specific objectives are: (1) to analyze the characteristics of family functioning among Chengdu’s primary and secondary school students; (2) To explore the pathways of association between family environmental factors—such as parent-child relationships, family support, and parenting styles—and psychological resilience and emotional competence. Existing research has confirmed positive associations between psychological resilience, family functioning, and emotional competence; however, in-depth studies focusing on primary and secondary school students remain limited. This group is in a critical period of social, emotional, and behavioral development, and their unique academic pressures and family environments may be closely related to the relationships among these variables. Moreover, while current studies predominantly focus on the direct link between psychological resilience and emotional competence, the mediating role of family functioning has yet to be fully explored. Through empirical analysis, this study aims to address this gap, refine the theoretical framework linking family environment and child psychological development, and provide theoretical and practical guidance for optimizing family education and enhancing students’ psychological well-being.

Method

Participants

The Chengdu Positive Child Development (CPCD) survey is a school-based cohort study conducted in Chengdu, the capital of Sichuan Province, which is home to 623 primary schools, 317 junior high schools, and 156 nine-year integrated schools. The baseline data for this study were collected from December 23, 2019, to January 13, 2020. The CPCD survey forms a key component of the Jockey Club Youth Enhancement Scheme’s “Tin Ka Ping Positive Youth Training Program.” To ensure sample representativeness, Chengdu was first stratified into different regions based on economic conditions. Within each stratum, one district was randomly selected using a random number generator, followed by the random selection of one school from the list of schools in each sampled district. Ultimately, five schools were chosen, including one primary school, one junior high school, and three nine-year integrated schools. Among the participating schools, one is located in the city center, two in the southern suburbs, and two in the northern suburbs. To comprehensively collect participant information, all willing children and adolescents from grades 1 to 9, along with their parents, were invited to participate, with no specific inclusion or exclusion criteria applied. Through stratified sampling, a total of 8,968 children and adolescents aged 6 to 16 were recruited from the five primary and junior high schools. Of the 8,124 questionnaires distributed, 7,937 valid responses were collected, yielding a response rate of 97.70%. The primary aim of this study is to assess the current state of positive child development, as well as psychosocial and behavioral issues, while evaluating the effectiveness of school-based health education and promotion programs in fostering positive youth development and addressing psychological and behavioral problems. Detailed descriptions of the study’s design and methodology have been provided in other literature [25].

Study design

This study employed a cross-sectional survey method, conducting self-administered questionnaires on a class-by-class basis. Prior to the survey, the research team explained the study’s objectives to the students and their legal guardians in detail, emphasizing data confidentiality and the voluntary nature of participation. Written informed consent was obtained from the legal guardians. The study received approval from the Medical Ethics Committee of ** University (Approval No. K2020025). The questionnaire was administered in a classroom setting, supervised by two trained research assistants to ensure that all students had sufficient time to complete it independently. For primary school students, particularly those in lower grades, the class teacher read each question aloud item-by-item and assisted in clarifying any doubts raised by the students. Older students read and answered the questions independently. Upon completion, the research assistants collected the questionnaires and conducted a quick review to ensure response completeness, requesting students to revise their answers if necessary. Detailed information on the study’s design and methodology can be found in the published cohort profile [25].

Measurement tools

General information questionnaire

Designed by the researchers, this questionnaire collects basic demographic information, including gender, age, ethnicity, grade level, and household registration location (urban/rural).

Chinese family assessment instrument (C-FAI)

This study utilized the Chinese Family Assessment Instrument (C-FAI) [26] to evaluate family functioning. Developed by Shek et al. in 2002, the C-FAI is designed to assess the degree of family dysfunction within the context of Chinese culture. The instrument comprises five dimensions: mutuality (12 items, mutual support, love and concern among family members, e.g., “Family members understand each other”), communication (9 items, frequency and nature of interaction among family members, e.g., “Family members are cohesive”), conflict and harmony (6 items, conflicting and harmonious behaviours in the family, e.g., “Poor marital relationship between parents”), parental concern (3 items, parentalsupport behaviours, e.g., “Parents take care of the children”), and parental control (3 items, harshness of parenting behaviours, e.g., “Parental control is too strict”). The C-FAI consists of 33 items in total, scored on a 5-point Likert scale ranging from 1 (“very similar”) to 5 (“very dissimilar”), with 9 items reverse-scored. The total score ranges from 33 to 165, with higher scores indicating greater family dysfunction. In this study, the C-FAI demonstrated strong reliability and validity, with a Cronbach’s α coefficient of 0.948, a Kaiser-Meyer-Olkin (KMO) value of 0.967 (close to 1), and a Bartlett’s Test of Sphericity yielding a significance level of < 0.001.

Resilience subscale (RE)

Students’ psychological resilience was measured using the “Resilience Subscale” from the “Chinese Positive Youth Development Scale” (CPYDS), developed and validated by Shek and Ma [27]. This subscale comprises 6 items: (A1) “When facing difficulties, I don’t give up easily”; (A2) “When confronted with adversity, I can maintain an optimistic mindset”; (A3) “I believe that people with ideals and determination will ultimately succeed”; (A4) “I believe: ‘Only by enduring the hardest hardships can one rise above others’”; (A5) “My belief is: ‘Even if tomorrow gets worse, I will still live well’”; (A6) “I believe that difficulties in life can be resolved.” The items are scored on a 6-point Likert scale, ranging from 1 (“strongly disagree”) to 6 (“strongly agree”), with total scores ranging from 6 to 36. Higher scores indicate greater psychological resilience. In this study, the analysis revealed a Cronbach’s α coefficient of 0.874, a Kaiser-Meyer-Olkin (KMO) value of 0.891, and a Bartlett’s Test of Sphericity significance level of < 0.001, demonstrating good reliability and validity of the questionnaire.

Emotional competence subscale (EC)

Students’ emotional competence was assessed using the “Emotional Competence Subscale” from the CPYDS, developed and validated by Shek [11]. This subscale consists of 6 items: (A1) “I am a cheerful person”; (A2) “If I am unhappy, I can appropriately express my emotions”; (A3) “If I get angry, I can usually express my feelings rationally”; (A4) “When in conflict with others, I can typically manage my emotions”; (A5) “I can see the world from others’ perspectives and understand their feelings”; (A6) “I let others know about my emotions.” The items are scored on a 6-point Likert scale, ranging from 1 (“strongly disagree”) to 6 (“strongly agree”), with total scores ranging from 6 to 36. Higher scores indicate better emotional control and expression abilities. In this study, the questionnaire demonstrated a Cronbach’s α coefficient of 0.866, a Kaiser-Meyer-Olkin (KMO) value of 0.878, and a Bartlett’s Test of Sphericity significance level of < 0.001, confirming its good reliability and validity.

Statistical analysis

This study utilized Epidata 3.1 software for double data entry and verification to ensure data accuracy and consistency, while data analysis was conducted using SPSS 26.0 software. For the basic characteristics of the participants, categorical data were described using frequencies and percentages, while continuous data were expressed as means ± standard deviations (M ± SD). All scales were subjected to the Harman single-factor test to assess common method bias. In examining differences in family functioning and students’ abilities, independent samples t-tests were used for comparing the means of two groups that met the assumptions of normality and homogeneity of variance; if variances were unequal, Welch’s t-test was applied. For comparisons of three or more groups satisfying normality and homogeneity of variance, one-way analysis of variance (ANOVA) was employed; if variances were unequal, Welch’s ANOVA was used. For data with unequal sample sizes and heterogeneous variances, post-hoc multiple comparisons were conducted using the Games-Howell method. Variables showing statistically significant differences in the aforementioned two-group comparisons were included in partial correlation analysis, controlling for other variables to examine correlations. Structural equation modeling was conducted using AMOS 26.0 software to analyze the associations and mediation effects among family functioning, psychological resilience, and emotional competence, with model fit evaluation (e.g., chi-square test, CFI, RMSEA) and adjustments. The large sample size (N = 7,937) enhanced the precision of parameter estimation and the stability of results, increasing the statistical power and generalizability of the study. Structural equation modeling (SEM) is widely used in cross-sectional studies to explore complex relationships and mediation effects among multiple variables, with its robustness effectively testing the association pathways among family functioning, psychological resilience, and emotional competence. The significance level for hypothesis testing was set at α = 0.05.

After initially constructing the Structural Equation Model (SEM), we found that the model’s fit indices did not meet the desired standards. By examining the model’s Modification Indices in conjunction with a content analysis of the scale items, we identified potential correlations among the residuals of some observed variables. This was primarily attributed to semantic or content overlap between certain items, reflecting specific nuances that were not fully captured by the latent variables. For instance, in the Emotional Competence scale, item EC2 (‘If I’m unhappy, I can express my emotions appropriately’) and item EC6 (‘I will let others know my emotions’) both specifically point to the dimension of ‘external emotional expression.’ In the Resilience scale, item A3 (‘I believe people with ideals and determination.will succeed’) and item A4 (‘I believe: “no pain, no gain"’) both reflect cultural beliefs about perseverance and reward. This high degree of content similarity provides the theoretical rationale for their residual correlation. Therefore, following established SEM practices that permit correlating residuals when there is a strong theoretical justification [28], we incorporated these specific residual covariance paths into our model. This modification was intended to account for the common method variance arising from item content overlap, thereby allowing the model to more accurately reflect the data structure and achieving a good model fit (CMIN/DF = 6.988, RMSEA = 0.027, CFI = 0.993).

Results

Basic characteristics of participants

This study included a total of 7,937 primary and secondary school students. Among them, 4,086 were boys (51.48%) and 3,851 were girls (48.52%); 99.18% were of Han ethnicity; 5,087 resided in urban areas (64.09%), while 2,850 were from rural areas (35.91%). There were 5,389 students in grades 1–6 (primary school stage, 68.01%), with an average age of 9.89 ± 0.21 years, and 2,539 students in grades 7–9 (early secondary school stage, 31.99%), with an average age of 13.44 ± 0.19 years.

Common method bias test

This study employed the Harman single-factor test to examine common method bias. The results revealed seven factors with eigenvalues greater than 1, with the first factor accounting for 34.0% of the variance, which did not exceed 50% of the total variance explained. This indicates that no single factor dominated the explanation of the variance, suggesting that common method bias was not a significant issue in the data [29].

Differences in family functioning, psychological resilience, and emotional competence across participant characteristics

The analysis showed significant differences in family functioning across gender, urban/rural residence, and grade level (P < 0.05) (Table 1). Specifically, boys scored higher in family functioning than girls (1.97 ± 0.74 vs. 1.88 ± 0.72). Across the five dimensions of family functioning, girls scored higher than boys in the dimension of mutual communication (mutual support, love, and concern among family members, and the frequency and nature of interactions), while boys scored higher in the dimensions of conflict and harmony (conflicting and harmonious behaviors in the family), parental concern (parental support behaviors), and parental control (harshness of parenting behaviors) (Table 2). Additionally, family functioning scores were generally higher in rural areas compared to urban areas (1.97 ± 0.73 vs. 1.90 ± 0.73) (Table 1). Multiple comparisons across the nine grade levels were conducted using the Games-Howell method (Table 3), revealing that children in grades 4–6 had significantly lower family functioning scores compared to other grades (P < 0.05). Within the grades 4–6 group, differences were minimal, while differences between other grade groups were not significant (P > 0.05).

Table 1.

Differences in dimensions of family functioning across basic characteristics of participants (N = 7937)

Variable	Category	Frequency	Percentage (100%)	Total Family Functioning Score			Total Resilience Score			Total Emotional Competence Score
				M ± SD	t/F	P	M ± SD	t/F	P	M ± SD	t/F	P
Gender	Girls	3851	48.52%	1.88 ± 0.72	24.643*	<0.001	31.55 ± 5.16	0.049	0.825	28.43 ± 6.58	0.371	0.542
	Boys	4086	51.48%	1.97 ± 0.74			31.52 ± 5.33			28.34 ± 6.77
Ethnicity	Han	7872	99.18%	1.93 ± 0.73	0.018	0.893	31.54 ± 5.25	1.550	0.213	28.39 ± 6.67	1.280	0.258
	Other	65	0.82%	1.91 ± 0.79			30.80 ± 5.78			27.35 ± 7.36
Population Type	Urban	5087	64.09%	1.90 ± 0.73	14.600	<0.001	31.62 ± 5.23	4.025	0.045	28.39 ± 6.75	0.018	0.892
	Rural	2850	35.91%	1.96 ± 0.73			31.38 ± 5.27			28.37 ± 6.55
Grade	1	357	4.50%	2.06 ± 0.68	13.921*	<0.001	31.93 ± 5.52	19.127*	<0.001	30.47 ± 6.49	32.140*	<0.001
	2	402	5.06%	2.00 ± 0.75			32.24 ± 4.58			29.19 ± 6.44
	3	1107	13.95%	1.95 ± 0.71			32.22 ± 4.15			28.01 ± 6.7
	4	1134	14.29%	1.83 ± 0.68			32.51 ± 4.03			29.13 ± 6.32
	5	1166	14.69%	1.83 ± 0.71			32.00 ± 4.89			28.32 ± 6.59
	6	1232	15.52%	1.85 ± 0.74			31.91 ± 5.61			28.99 ± 6.93
	7	1107	13.95%	1.99 ± 0.79			30.75 ± 6.24			28.36 ± 6.97
	8	1079	13.59%	2.04 ± 0.73			29.64 ± 5.90			26.82 ± 6.49
	9	353	4.45%	2.01 ± 0.73			30.44 ± 4.66			27.03 ± 5.89

* Indicates unequal variance, analyzed using the Welch test

Table 2.

Differences in family functioning and total scores of various abilities across basic characteristics of participants (N = 7937)

Variable	Category	Mutuality (Mut)			Communication (Com)			Conflict and Harmony (CH)			Parental Concern (PCC)			Parental Control (PCT)
		M ± SD	t/F	P	M ± SD	t/F	P	M ± SD	t/F	P	M ± SD	t/F	P	M ± SD	t/F	P
Gender	Girls	1.69 ± 0.83	0.892	0.345	1.88 ± 0.95	11.076*	0.001	2.06 ± 0.90	52.265*	<0.001	1.61 ± 0.86	35.741*	<0.001	2.19 ± 1.19	63.443*	<0.001
	Boys	1.67 ± 0.84			1.81 ± 0.93			2.22 ± 0.99			1.72 ± 0.91			2.42 ± 1.30
Population Type	Urban	1.66 ± 0.84	6.208	0.013	1.81 ± 0.94	18.319	<0.001	2.12 ± 0.95	7.392*	0.007	1.65 ± 0.88	8.183	0.004	2.28 ± 1.26	4.844	0.028
	Rural	1.71 ± 0.83			1.90 ± 0.95			2.18 ± 0.94			1.71 ± 0.91			2.35 ± 1.25
Grade	1	1.47 ± 0.72	15.468*	<0.001	1.53 ± 0.80	24.205*	<0.001	2.61 ± 1.01	23.581*	<0.001	1.90 ± 0.91	12.655*	<0.001	2.78 ± 1.47	8.062*	<0.001
	2	1.65 ± 0.81			1.71 ± 0.84			2.39 ± 1.00			1.81 ± 0.98			2.44 ± 1.38
	3	1.65 ± 0.81			1.80 ± 0.90			2.25 ± 0.92			1.68 ± 0.94			2.37 ± 1.33
	4	1.59 ± 0.76			1.70 ± 0.87			2.05 ± 0.92			1.57 ± 0.90			2.23 ± 1.23
	5	1.62 ± 0.82			1.79 ± 0.92			2.01 ± 0.90			1.55 ± 0.83			2.20 ± 1.21
	6	1.61 ± 0.84			1.82 ± 0.96			1.98 ± 0.95			1.59 ± 0.86			2.22 ± 1.23
	7	1.79 ± 0.93			1.95 ± 1.03			2.16 ± 0.96			1.72 ± 0.90			2.32 ± 1.25
	8	1.83 ± 0.83			2.05 ± 0.94			2.18 ± 0.95			1.78 ± 0.85			2.35 ± 1.17
	9	1.85 ± 0.84			2.13 ± 0.98			2.10 ± 0.88			1.70 ± 0.81			2.25 ± 1.13

* Indicates unequal variance, analyzed using the Welch test

Table 3.

Games-Howell pairwise comparison results of family functioning scores across different grade levels (N = 7937)

(I)group	(J)group	MD(I-J)	SE	Sig.	95%CI LowerBound	95%CI UpperBound
1	2	0.057	0.052	0.975	−0.105	0.219
	3	0.111	0.042	0.170	−0.020	0.242
	4	0.230*	0.041	0.000	0.101	0.359
	5	0.224*	0.042	0.000	0.094	0.354
	6	0.213*	0.042	0.000	0.082	0.343
	7	0.072	0.043	0.767	−0.063	0.207
	8	0.019	0.042	1.000	−0.113	0.151
	9	0.052	0.053	0.987	−0.113	0.218
2	1	−0.057	0.052	0.975	−0.219	0.105
	3	0.054	0.043	0.945	−0.081	0.188
	4	0.172*	0.043	0.002	0.040	0.305
	5	0.167*	0.043	0.004	0.033	0.301
	6	0.155*	0.043	0.010	0.022	0.289
	7	0.015	0.044	1.000	−0.123	0.153
	8	−0.038	0.044	0.994	−0.174	0.098
	9	−0.005	0.054	1.000	−0.173	0.164
3	1	−0.111	0.042	0.170	−0.242	0.020
	2	−0.054	0.043	0.945	−0.188	0.081
	4	0.119*	0.029	0.002	0.027	0.210
	5	0.113*	0.030	0.005	0.020	0.206
	6	0.101*	0.030	0.021	0.008	0.194
	7	−0.039	0.032	0.953	−0.138	0.060
	8	−0.092	0.031	0.071	−0.188	0.004
	9	−0.059	0.044	0.925	−0.197	0.080
4	1	−0.230*	0.041	0.000	−0.359	−0.101
	2	−0.172*	0.043	0.002	−0.305	−0.040
	3	−0.119*	0.029	0.002	−0.210	−0.027
	5	−0.006	0.029	1.000	−0.096	0.085
	6	−0.017	0.029	1.000	−0.108	0.073
	7	−0.157*	0.031	0.000	−0.255	−0.061
	8	−0.211*	0.030	0.000	−0.304	−0.117
	9	−0.177*	0.044	0.002	−0.314	−0.041
5	1	−0.224*	0.042	0.000	−0.354	−0.094
	2	−0.167*	0.043	0.004	−0.301	−0.033
	3	−0.113*	0.030	0.005	−0.206	−0.020
	4	0.006	0.029	1.000	−0.085	0.096
	6	−0.011	0.030	1.000	−0.103	0.081
	7	−0.152*	0.032	0.000	−0.250	−0.053
	8	−0.205*	0.031	0.000	−0.300	−0.110
	9	−0.171*	0.044	0.004	−0.309	−0.034
6	1	−0.213*	0.042	0.000	−0.343	−0.082
	2	−0.155*	0.043	0.010	−0.289	−0.022
	3	−0.101*	0.030	0.021	−0.194	−0.008
	4	0.017	0.029	1.000	−0.073	0.108
	5	0.011	0.030	1.000	−0.081	0.103
	7	−0.140*	0.032	0.000	−0.239	−0.042
	8	−0.193*	0.031	0.000	−0.288	−0.098
	9	−0.160*	0.044	0.010	−0.298	−0.022
7	1	−0.072	0.043	0.767	−0.207	0.063
	2	−0.015	0.044	1.000	−0.153	0.123
	3	0.039	0.032	0.953	−0.060	0.138
	4	0.157*	0.031	0.000	0.061	0.255
	5	0.152*	0.032	0.000	0.053	0.250
	6	0.140*	0.032	0.000	0.042	0.239
	8	−0.053	0.033	0.790	−0.154	0.048
	9	−0.020	0.046	1.000	−0.162	0.122
8	1	−0.019	0.042	1.000	−0.151	0.113
	2	0.038	0.044	0.994	−0.098	0.174
	3	0.092	0.031	0.071	−0.004	0.188
	4	0.211*	0.030	0.000	0.117	0.304
	5	0.205*	0.031	0.000	0.110	0.300
	6	0.193*	0.031	0.000	0.098	0.288
	7	0.053	0.033	0.790	−0.048	0.154
	9	0.033	0.045	0.998	−0.106	0.173
9	1	−0.052	0.053	0.987	−0.218	0.113
	2	0.005	0.054	1.000	−0.164	0.173
	3	0.059	0.044	0.925	−0.080	0.197
	4	0.177*	0.044	0.002	0.041	0.314
	5	0.171*	0.044	0.004	0.034	0.309
	6	0.160*	0.044	0.010	0.022	0.298
	7	0.020	0.046	1.000	−0.122	0.162
	8	−0.033	0.045	0.998	−0.173	0.106

*Significance level of the mean difference set at 0.05

Regarding psychological resilience, the Games-Howell analysis results (Table 4) indicated that students in grades 7–9 had lower resilience scores compared to other grades (P < 0.05), with children in grades 1–6 exhibiting the highest resilience scores. Differences within the grades 1–6 group were not significant (P > 0.05).

Table 4.

Games-Howell pairwise comparison results of psychological resilience scores across different grade levels (N = 7937)

(I)group	(J)group	MD(I-J)	SE	Sig.	95%CI LowerBound	95%CI UpperBound
1	2	−0.317	0.371	0.995	−1.470	0.837
	3	−0.294	0.318	0.991	−1.284	0.695
	4	−0.585	0.316	0.646	−1.569	0.398
	5	−0.069	0.325	1.000	−1.082	0.945
	6	0.022	0.333	1.000	−1.015	1.059
	7	1.176*	0.347	0.021	0.096	2.257
	8	2.286*	0.343	0.000	1.218	3.353
	9	1.482*	0.383	0.004	0.290	2.675
2	1	0.317	0.371	0.995	−0.837	1.470
	3	0.022	0.260	1.000	−0.787	0.832
	4	−0.269	0.258	0.982	−1.071	0.534
	5	0.248	0.270	0.992	−0.591	1.087
	6	0.339	0.279	0.953	−0.528	1.205
	7	1.493*	0.295	0.000	0.575	2.411
	8	2.602*	0.291	0.000	1.699	3.506
	9	1.799*	0.337	0.000	0.751	2.847
3	1	0.294	0.318	0.991	−0.695	1.284
	2	−0.022	0.260	1.000	−0.832	0.787
	4	−0.291	0.173	0.756	−0.827	0.245
	5	0.226	0.190	0.959	−0.364	0.815
	6	0.316	0.203	0.826	−0.313	0.945
	7	1.471*	0.225	0.000	0.772	2.170
	8	2.580*	0.219	0.000	1.901	3.259
	9	1.777*	0.277	0.000	0.913	2.641
4	1	0.585	0.316	0.646	−0.398	1.569
	2	0.269	0.258	0.982	−0.534	1.071
	3	0.291	0.173	0.756	−0.245	0.827
	5	0.517	0.187	0.126	−0.063	1.096
	6	0.607	0.200	0.060	−0.012	1.227
	7	1.762*	0.222	0.000	1.071	2.452
	8	2.871*	0.216	0.000	2.201	3.542
	9	2.068*	0.275	0.000	1.210	2.925
5	1	0.069	0.325	1.000	−0.945	1.082
	2	−0.248	0.270	0.992	−1.087	0.591
	3	−0.226	0.190	0.959	−0.815	0.364
	4	−0.517	0.187	0.126	−1.096	0.063
	6	0.091	0.215	1.000	−0.576	0.757
	7	1.245*	0.236	0.000	0.513	1.978
	8	2.354*	0.230	0.000	1.641	3.068
	9	1.551*	0.286	0.000	0.660	2.442
6	1	−0.022	0.333	1.000	−1.059	1.015
	2	−0.339	0.279	0.953	−1.205	0.528
	3	−0.316	0.203	0.826	−0.945	0.313
	4	−0.607	0.200	0.060	−1.227	0.012
	5	−0.091	0.215	1.000	−0.757	0.576
	7	1.154*	0.246	0.000	0.390	1.919
	8	2.264*	0.240	0.000	1.517	3.010
	9	1.460*	0.295	0.000	0.543	2.378
7	1	−1.176*	0.347	0.021	−2.257	−0.096
	2	−1.493*	0.295	0.000	−2.411	−0.575
	3	−1.471*	0.225	0.000	−2.170	−0.772
	4	−1.762*	0.222	0.000	−2.452	−1.071
	5	−1.245*	0.236	0.000	−1.978	−0.513
	6	−1.154*	0.246	0.000	−1.919	−0.390
	8	1.109*	0.260	0.001	0.303	1.916
	9	0.306	0.311	0.987	−0.661	1.273
8	1	−2.286*	0.343	0.000	−3.353	−1.218
	2	−2.602*	0.291	0.000	−3.506	−1.699
	3	−2.580*	0.219	0.000	−3.259	−1.901
	4	−2.871*	0.216	0.000	−3.542	−2.201
	5	−2.354*	0.230	0.000	−3.068	−1.641
	6	−2.264*	0.240	0.000	−3.010	−1.517
	7	−1.109*	0.260	0.001	−1.916	−0.303
	9	−0.803	0.306	0.178	−1.756	0.149
9	1	−1.482*	0.383	0.004	−2.675	−0.290
	2	−1.799*	0.337	0.000	−2.847	−0.751
	3	−1.777*	0.277	0.000	−2.641	−0.913
	4	−2.068*	0.275	0.000	−2.925	−1.210
	5	−1.551*	0.286	0.000	−2.442	−0.660
	6	−1.460*	0.295	0.000	−2.378	−0.543
	7	−0.306	0.311	0.987	−1.273	0.661
	8	0.803	0.306	0.178	−0.149	1.756

*Significance level of the mean difference set at 0.05

In terms of emotional competence (Table 5), children in grade 1 scored the highest, while adolescents in grades 8–9 scored the lowest; however, these differences were not statistically significant (P > 0.05).

Table 5.

Games-Howell pairwise comparison results of emotional competence scores across different grade levels (N = 7937)

(I)group	(J)group	MD(I-J)	SE	Sig.	95%CI LowerBound	95%CI UpperBound
1	2	1.279	0.470	0.142	−0.184	2.742
	3	2.463*	0.398	0.000	1.224	3.703
	4	1.342*	0.392	0.019	0.123	2.561
	5	2.156*	0.394	0.000	0.929	3.383
	6	1.487*	0.396	0.006	0.254	2.721
	7	2.118*	0.403	0.000	0.865	3.371
	8	3.650*	0.396	0.000	2.417	4.884
	9	3.448*	0.465	0.000	2.001	4.895
2	1	−1.279	0.470	0.142	−2.742	0.184
	3	1.184*	0.379	0.048	0.005	2.363
	4	0.063	0.372	1.000	−1.094	1.220
	5	0.877	0.375	0.319	−0.289	2.042
	6	0.208	0.377	1.000	−0.964	1.380
	7	0.839	0.383	0.414	−0.354	2.032
	8	2.371*	0.377	0.000	1.199	3.544
	9	2.169*	0.449	0.000	0.773	3.564
3	1	−2.463*	0.398	0.000	−3.703	−1.224
	2	−1.184*	0.379	0.048	−2.363	−0.005
	4	−1.121*	0.275	0.002	−1.976	−0.267
	5	−0.307	0.279	0.974	−1.173	0.559
	6	−0.976*	0.282	0.016	−1.852	−0.101
	7	−0.345	0.291	0.959	−1.248	0.558
	8	1.187*	0.282	0.001	0.311	2.063
	9	0.984	0.372	0.171	−0.175	2.144
4	1	−1.342*	0.392	0.019	−2.561	−0.123
	2	−0.063	0.372	1.000	−1.220	1.094
	3	1.121*	0.275	0.002	0.267	1.976
	5	0.814	0.269	0.063	−0.022	1.650
	6	0.145	0.272	1.000	−0.700	0.991
	7	0.776	0.281	0.128	−0.097	1.650
	8	2.308*	0.272	0.000	1.463	3.154
	9	2.106*	0.365	0.000	0.969	3.243
5	1	−2.156*	0.394	0.000	−3.383	−0.929
	2	−0.877	0.375	0.319	−2.042	0.289
	3	0.307	0.279	0.974	−0.559	1.173
	4	−0.814	0.269	0.063	−1.650	0.022
	6	−0.669	0.276	0.271	−1.526	0.188
	7	−0.038	0.285	1.000	−0.922	0.847
	8	1.494*	0.276	0.000	0.637	2.352
	9	1.292*	0.368	0.014	0.147	2.437
6	1	−1.487*	0.396	0.006	−2.721	−0.254
	2	−0.208	0.377	1.000	−1.380	0.964
	3	0.976*	0.282	0.016	0.101	1.852
	4	−0.145	0.272	1.000	−0.991	0.700
	5	0.669	0.276	0.271	−0.188	1.526
	7	0.631	0.288	0.411	−0.263	1.525
	8	2.163*	0.279	0.000	1.297	3.030
	9	1.961*	0.370	0.000	0.808	3.113
7	1	−2.118*	0.403	0.000	−3.371	−0.865
	2	−0.839	0.383	0.414	−2.032	0.354
	3	0.345	0.291	0.959	−0.558	1.248
	4	−0.776	0.281	0.128	−1.650	0.097
	5	0.038	0.285	1.000	−0.847	0.922
	6	−0.631	0.288	0.411	−1.525	0.263
	8	1.532*	0.288	0.000	0.638	2.426
	9	1.330*	0.377	0.013	0.157	2.503
8	1	−3.650*	0.396	0.000	−4.884	−2.417
	2	−2.371*	0.377	0.000	−3.544	−1.199
	3	−1.187*	0.282	0.001	−2.063	−0.311
	4	−2.308*	0.272	0.000	−3.154	−1.463
	5	−1.494*	0.276	0.000	−2.352	−0.637
	6	−2.163*	0.279	0.000	−3.030	−1.297
	7	−1.532*	0.288	0.000	−2.426	−0.638
	9	−0.203	0.370	1.000	−1.355	0.950
9	1	−3.448*	0.465	0.000	−4.895	−2.001
	2	−2.169*	0.449	0.000	−3.564	−0.773
	3	−0.984	0.372	0.171	−2.144	0.175
	4	−2.106*	0.365	0.000	−3.243	−0.969
	5	−1.292*	0.368	0.014	−2.437	−0.147
	6	−1.961*	0.370	0.000	−3.113	−0.808
	7	−1.330*	0.377	0.013	−2.503	−0.157
	8	0.203*	0.370	1.000	−0.950	1.355

*Significance level of the mean difference set at 0.05

Correlation analysis of family functioning with psychological resilience and emotional competence

The survey results revealed that, after controlling for variables such as gender, urban/rural residence, and grade level, the total family functioning score was significantly negatively correlated with both the total psychological resilience score (r = −0.394, P < 0.001) and the total emotional competence score (r = −0.376, P < 0.001). Similarly, the dimensions of family functioning also show a negative correlation with both the total score of resilience and emotional competence. Additionally, a significant positive correlation was observed between the total psychological resilience score and emotional competence (r = 0.646, P < 0.001) (Table 6).

Table 6.

Correlation analysis of family functioning, psychological resilience, and emotional regulation and expression abilities (N = 7937)

Dimension	M ± SD	FFT	Mut	Com	CH	PC	PCT	RE	RC
FFT: Total Family Dysfunction Score	1.93 ± 0.73	1
Mut: Mutuality	1.68 ± 0.83	0.764***	1
Com; communication	1.84 ± 0.94	0.742***	0.846***	1
CH: conflict and harmony	2.14 ± 0.95	0.759***	0.373***	0.314***	1
PC: parental concern	1.67 ± 0.89	0.803***	0.592***	0.546***	0.529***	1
PCT: parental control	2.31 ± 1.26	0.720***	0.237***	0.237***	0.607***	0.439***	1
RE: Total Psychological Resilience Score	31.53 ± 5.25	−0.394***	−0.354***	−0.357***	−0.273***	−0.300***	−0.231***	1
EC: Total Emotional Competence Score	28.38 ± 6.68	−0.376***	−0.372***	−0.405***	−0.228***	−0.256***	−0.199***	0.646***	1

*** P < 0.001 (two-tailed)

Mediation effect analysis of family functioning between psychological resilience and emotional competence

Using AMOS 26.0, a structural equation model was constructed for mediating effect analysis, and a multi-group structural equation model was employed to test the measurement invariance of the evaluation tool across different age groups in elementary and secondary schools. Based on the changes in the fit indices, the results show that there are no significant differences in measurement invariance between the elementary and secondary school groups, and the effect of grade grouping on model fit is minimal. Therefore, it can be concluded that grade has no significant impact on the model, with the detailed results presented in Tables 7, 8 and 9. The model fit indices were as follows: CMIN/DF (chi-square to degrees of freedom ratio) = 6.988, RMSEA (root mean square error of approximation) = 0.027, TLI (Tucker-Lewis Index) = 0.988, CFI (comparative fit index) = 0.993, and SRMR (standardized root mean square residual) < 0.001. These results indicate an overall good model fit (Fig. 2).

Table 7.

Model fit indices for unconstrained model in Multi-group SEM analysis (Elementary vs. Secondary School)

Model	DF	CMIN	P	NFI Delta-1	IFI Delta-2	RFI rho-1	TLI rho2
Measurement weights	14	78.740	0.000	0.001	0.001	0.000	0.000
Structural weights	17	157.725	0.000	0.002	0.002	0.001	0.001
Structural covariances	18	289.928	0.000	0.004	0.004	0.004	0.004
Structural residuals	20	335.488	0.000	0.005	0.005	0.005	0.005
Measurement residuals	37	1072.145	0.000	0.015	0.016	0.016	0.016

Table 8.

Model fit indices for measurement weights constrained in Multi-group SEM analysis (Elementary vs. Secondary School)

Model	DF	CMIN	P	NFI Delta-1	IFI Delta-2	RFI rho-1	TLI rho2
Structural weights	3	78.986	0.000	0.001	0.001	0.001	0.001
Structural covariances	4	211.188	0.000	0.003	0.003	0.004	0.004
Structural residuals	6	256.748	0.000	0.004	0.004	0.005	0.005
Measurement residuals	23	993.405	0.000	0.014	0.014	0.016	0.016

Table 9.

Model fit indices for structural weights constrained in Multi-group SEM analysis (Elementary vs. Secondary School)

Model	DF	CMIN	P	NFI Delta-1	IFI Delta-2	RFI rho-1	TLI rho2
Structural covariances	1	132.203	0.000	0.002	0.002	0.003	0.003
Structural residuals	3	177.763	0.000	0.003	0.003	0.003	0.003
Measurement residuals	20	914.420	0.000	0.013	0.013	0.015	0.015

Fig. 2.

Mediation Effect Analysis of Family Functioning Between Psychological Resilience and Emotional Competence

The results of the mediation analysis showed that the main standardized path coefficients of the model were all significant (P < 0.001): the standardized path coefficient from psychological resilience to family functioning was − 0.43, the standardized path coefficient from family functioning to emotional competence was − 0.16, and the standardized direct path coefficient from psychological resilience to emotional competence was 0.70 (see Table 10 for specific unstandardized coefficients and model parameters).

Table 10.

The estimated path coefficients for the model of family functioning, resilience, and emotional competence (N = 7937)

		Estimate	S.E.	C.R.	P	Label
FAI	<---	RE	− 0.364	0.011	−31.671	***
EC	<---	FAI	− 0.243	0.015	−16.249	***
EC	<---	RE	0.871	0.020	43.622	***
Mut_x	<---	FAI	1.000
Com_x	<---	FAI	1.219	0.018	69.311	***
CH_x	<---	FAI	0.539	0.017	32.667	***
PC_x	<---	FAI	0.722	0.022	32.188	***
PCT_x	<---	FAI	0.699	0.037	19.090	***
RE6	<---	RE	1.000
RE5	<---	RE	0.976	0.018	54.491	***
RE4	<---	RE	0.758	0.016	46.758	***
RE3	<---	RE	0.887	0.017	51.667	***
RE2	<---	RE	1.123	0.020	54.990	***
RE1	<---	RE	0.853	0.017	51.279	***
EC1	<---	EC	1.000
EC2	<---	EC	0.948	0.016	58.242	***
EC3	<---	EC	0.986	0.019	52.181	***
EC4	<---	EC	0.924	0.016	57.563	***
EC5	<---	EC	0.812	0.015	52.547	***
EC6	<---	EC	0.933	0.020	47.664	***

*** P < 0.001

Furthermore, a bootstrap test with 5,000 iterations revealed that the standardized mediation effect of psychological resilience on emotional competence through family functioning was 0.089, with a 95% confidence interval of 0.667 to 0.726, excluding 0. This suggests a significant mediation effect, accounting for 9.2% of the total effect. Detailed results are presented in Tables 10 and 11.

Table 11.

Decomposition table of the mediation effect of family functioning between psychological resilience and emotional competence (N = 7937)

Mediation Pathway	Standardized Effect Size	95%CI Lower Limit	95%CI Upper Limit	P	Proportion of Effect
Total Effect	0.960	0.909	1.013	<0.001
Direct Effect	0.871	0.815	0.928	<0.001
Mediation Effect	0.089	0.074	0.105	<0.001	9.2%

Discussion

Current Status of Family Functioning, Psychological Resilience, and Emotional Competence Among Primary and Secondary School Students in Chengdu

This study identified significant differences in family functioning across gender, urban/rural residence, and grade level, consistent with the dynamic nature of family functioning as a social support system [30]. Firstly, girls scored higher than boys in the dimension of mutual family communication, while boys scored higher in the dimensions of conflict and harmony, parental concern, and parental control. This gender difference may be associated with social gender role expectations: traditional norms often encourage girls to develop emotional expression and communication skills, whereas boys are frequently subjected to greater demands related to family rule enforcement and conflict resolution. Sun and Stewart’s research [31] found that Australian primary school girls outperformed boys in individual traits such as communication and empathy, aligning with the higher scores of girls in the family communication dimension observed in this study, suggesting that gender may be related to family interaction patterns. However, Nelson et al. [32] reported that girls perceived higher levels of parental monitoring and family conflict compared to boys, which may be associated with stricter normative expectations for girls. This study did not further analyze gender as a moderating variable; future research could use more complex models to examine the associations between gender-specific parenting styles and family functioning.

Regarding urban-rural differences, this study found that family functioning scores were significantly higher in rural areas compared to urban areas, this suggests that rural families may experience greater family dysfunction. This phenomenon may be associated with higher economic pressures, a higher proportion of parents working away from home, and reduced parent-child interactions commonly faced by rural families [33]. Long-term parental separation may be related to negative psychological adaptation among left-behind rural children, particularly in situational coping and cooperative adaptation abilities [34]. Such parental separation may be associated with dysfunction in emotional support and family interactions in rural families, consistent with prior research on urban-rural family environment differences [35]. Future research could further explore the mechanisms by which urban-rural differences moderate the relationship between family functioning and mental health.

Regarding grade-level differences, children in grades 4–6 had lower family functioning scores (indicating less dysfunction), which may be associated with their greater dependence on family at this stage. At this age, parents tend to provide more emotional and material support, family conflicts are relatively infrequent, and the family support system may be more stable. However, as students progress to grades 7–9 and enter adolescence, their heightened self-awareness may be associated with increased ideological conflicts with parents, which may, in turn, be accompanied by a greater degree of family dysfunction [36–38]. Future studies could further investigate the dynamic mechanisms underlying changes in family functioning with age to better understand the interplay between grade level and family functioning.

The results of this study also indicate that psychological resilience scores were significantly lower among students in grades 7–9 compared to other grades (P < 0.05), while children in grades 1–6 scored the highest, with no significant within-group differences (P > 0.05). This finding may be associated with the adaptive challenges faced by adolescents in grades 7–9 during their developmental process. Existing research suggests that junior high school students often exhibit lower psychological resilience due to intensified academic pressure and increased parent-child conflicts [39]. As grade levels rise, adolescents must navigate higher academic expectations and more complex social relationships, which may undermine their ability to cope with setbacks. In contrast, children in grades 1–6, during the primary school stage, typically benefit from stable family support and lower external pressures, resulting in higher psychological resilience [40]. The resilience scores of primary school children showed no significant differences across grades, which may be associated with the similarity of their developmental tasks and the stability of their external environment. This finding suggests that the junior high school stage may be a critical period for the development of psychological resilience, and future research could further validate the role of grade level as a moderating variable.

Additionally, the study found that grade 1 children scored the highest in emotional competence, while students in grades 8–9 scored the lowest, though the differences across grades were not statistically significant (P > 0.05). The preschool period is a critical stage for emotional competence development, laying the foundation for children’s adaptation after entering school. Preschoolers who can remain calm during conflicts often exhibit stronger social adjustment and academic performance in grade 1, and their high emotional competence scores may stem from systematic early education focused on basic emotion recognition and expression [41]. In early school years, children, guided by parents and teachers, tend to effectively understand and express simple emotions, which may enhance their performance on emotional competence assessments. However, the lower scores observed among students in grades 8–9 may be linked to the increased complexity of emotional regulation demands during adolescence [38]. Late junior high school adolescents face more complex emotional situations (e.g., shame or guilt) and dual pressures from academics and peer relationships, which may be associated with suppressed emotional competence performance. Although the differences across grades were not statistically significant, this trend suggests that the cultivation of emotional competence holds particular significance in late junior high school. Future research could further validate the applicability of measurement tools across different age groups or employ longitudinal designs to explore the underlying mechanisms of emotional competence changes with grade progression.

Correlation analysis of family functioning with psychological resilience and emotional competence

This study utilized the Family Functioning Assessment Scale (encompassing five dimensions: mutuality, communication, conflict and harmony, parental concern, and parental control) to examine the relationships between family functioning, psychological resilience, and emotional competence. The results revealed a significant negative correlation between the total family functioning score and both the total psychological resilience score (r = −0.394, P < 0.001) and the total emotional competence score (r = −0.376, P < 0.001). Concurrently, a significant positive correlation was observed between the total psychological resilience score and emotional competence (r = 0.646, P < 0.001).

First, the negative correlation between the total family functioning score and psychological resilience and emotional competence suggests that family dysfunction may be associated with risks to adolescent psychological development. Specifically, the five dimensions of family functioning may be related to psychological resilience and emotional competence through different mechanisms. The mutuality dimension reflects the support and emotional bonding among family members. Lower mutuality may lead adolescents to perceive a lack of family support, thereby weakening their psychological resilience [40]. Therefore, children in grades 1–6 may exhibit higher resilience due to stronger family support, while students in grades 7–9 may be associated with lower resilience due to decreased mutuality (e.g., estranged parent-child relationships). The communication dimension is related to emotional exchanges among family members. Inadequate communication may hinder adolescents from learning emotion expression and regulation skills, thus reducing their emotional competence [42]. The conflict and harmony dimension is directly related to the level of tension within the family. Frequent family conflicts may exacerbate emotional fluctuations in adolescents, diminishing their ability to cope with setbacks, particularly during the adolescent phase of grades 7–9 [36]. The parental concern dimension reflects the quality of parental support. Insufficient parental care may be associated with adolescents’ lack of security when facing stress, which may, in turn, be related to reduced psychological resilience. The parental control dimension, through excessive control, may restrict adolescent autonomy, triggering parent-child conflicts and subsequently undermining both emotional competence and psychological resilience [39]. Collectively, the manifestations of family dysfunction across these dimensions (e.g., low mutuality, high conflict) may be associated with restrictions in adolescent psychological development.

Second, the significant positive correlation between psychological resilience and emotional competence (r = 0.646, P < 0.001) suggests that the two may be closely associated in adolescent development. Individuals with high psychological resilience typically exhibit stronger stress-coping abilities, which may manifest through effective emotional regulation, thereby enhancing emotional competence [40]. Children in grades 1–6 may exhibit higher psychological resilience due to stable family support (e.g., high mutuality and parental care), which may provide a foundation for the cultivation of emotional competence. In contrast, students in grades 7–9 may be associated with a downward trend in both resilience and emotional competence due to family dysfunction (e.g., increased conflict and excessive parental control) [38]. The strength of this positive correlation (r > 0.6) suggests that psychological resilience may be related to an important association with emotional competence, warranting further validation in future research.

However, the negative or positive nature of these correlations does not directly establish causality. Family dysfunction may be associated with declines in psychological resilience and emotional competence, but adolescent emotional problems or low resilience may also be related to increased family conflicts, forming a bidirectional association [43]. In this study, the five dimensions of family functioning (mutuality, communication, conflict and harmony, parental care, and parental control) showed similar negative correlations with the total scores of psychological resilience and emotional competence, suggesting that the overall associations of these dimensions with adolescent psychological development are relatively consistent. Therefore, independent mediation analyses for each dimension were not conducted to avoid the limited contribution of repetitive analyses. However, considering that these dimensions theoretically have unique mechanisms (e.g., excessive parental control may be associated with emotional competence through different pathways), future research could further explore the independent contributions and potential mediation effects of each dimension using structural equation modeling (SEM) to more precisely reveal their differential associations. From a practical perspective, the study results highlight the positive significance of optimizing family functioning in relation to adolescent psychological health development. For the mutuality and communication dimensions, family intervention programs (e.g., parent-child communication training) could strengthen family support. For the conflict and harmony dimension, family mediation strategies could be introduced to reduce tensions. For the parental concern and control dimensions, parents should be encouraged to balance support and control to avoid overreach. Given the positive correlation between psychological resilience and emotional competence, schools could implement integrated intervention programs, such as emotional management courses and resilience training, to help adolescents navigate developmental challenges effectively.

Mediation effect analysis of family functioning between psychological resilience and emotional competence

The results of the mediation analysis further revealed the association pathways among psychological resilience, family functioning, and emotional competence. The analysis primarily revealed the following points:

1. Path a (Psychological Resilience → Family Functioning): Psychological resilience was significantly and negatively correlated with family dysfunction (β=−0.431,P < 0.001), meaning that adolescents with higher levels of psychological resilience reported lower degrees of family dysfunction. This suggests that individuals with stronger psychological resilience may be more capable of using positive coping strategies to improve family interactions and mitigate conflicts, thereby promoting healthier family functioning [40].

2. Path b (Family Functioning → Emotional Competence): Family dysfunction was significantly and negatively correlated with emotional competence (β=−0.165,P < 0.001), indicating that healthier family functioning (fewer dysfunctions) is associated with higher levels of emotional competence. This result is consistent with the finding in this study that all five dimensions of family functioning are correlated with emotional competence, highlighting the importance of a stable family environment in nurturing adolescent emotional competence.

3. Direct Effect c’ (Psychological Resilience → Emotional Competence): After controlling for the mediating effect of family functioning, the direct positive association of psychological resilience with emotional competence remained highly significant (β = 0.698,P < 0.001). This indicates that psychological resilience is a core internal trait that can independently and robustly associate with emotional competence. This strong direct association may stem from shared underlying mechanisms such as emotion regulation [38].

4. Indirect Effect (Mediating Effect): The Bootstrap test results confirmed that family functioning plays a significant partial mediating role between psychological resilience and emotional competence, with the indirect effect accounting for 9.2% of the total effect.

This finding reveals an important dual-path mechanism: psychological resilience is not only directly and strongly associated with emotional competence but also exerts its influence through the indirect pathway of “enhancing psychological resilience → improving family functioning → promoting emotional competence.” Although this is an indirect path with a moderate effect size, it clearly demonstrates that the family environment plays a key bridging role in the process by which adolescents transform internal psychological resources (psychological resilience) into practical emotional skills (emotional competence).

Although the mediation effect (9.2%) is relatively small, in psychological research, the effect size of a single mediator is typically limited, particularly for variables like emotional competence influenced by multiple factors [44]. This preliminary finding provides evidence for the mediating role of family functioning in a large-sample context, with particular practical significance for public health and educational interventions. Qiao et al. reported that family functioning is associated with problem behaviors in preschool left-behind children through a chained mediation of emotion regulation and psychological resilience, with a mediation effect accounting for 7.72% of the total effect, highly comparable to the 9.2% in this study, underscoring the associated role of family functioning in children’s emotional and behavioral development [45]. This suggests that family functioning, as an external support system, may be closely related to the development of emotional competence in primary and secondary school students [1, 10].

Specifically, the five dimensions of family functioning may be associated with emotional competence through different mechanisms in the mediation process: higher mutuality and parental care may be related to enhanced parent-child bonding, which may, in turn, be associated with improved emotional competence, consistent with the positive role of adaptive parent-child relationships [23]; conversely, family conflict or overly controlling parenting styles may be associated with increased communication barriers or parent-child tension, which may, in turn, be related to a weakened association between psychological resilience and emotional competence, particularly pronounced in father-adolescent relationships [46]. This mediation effect may be particularly relevant in the context of adolescence (grades 7–9), as parent-child conflicts and family dysfunction may be associated with heightened adolescent self-awareness [36, 37]. From a practical perspective, the 9.2% mediation effect suggests that optimizing family functioning may be associated with enhanced student emotional competence. School and family interventions can facilitate the improvement of emotional competence by enhancing family communication and support. Future research could explore other mediators (e.g., social support) or employ longitudinal designs to further validate the mediating role of family functioning and its generalizability across different cultural contexts.

Intervention design and application in educational practice

Based on the 9.2% mediation effect identified in this large-scale cross-sectional study (N = 7,937), the intervention design aims to optimize family functioning to support psychological resilience and emotional competence, providing preliminary guidance for educational practice.

1. Addressing the Negative Relationship Between Psychological Resilience and Family Functioning: Interventions aimed at enhancing adolescent resilience can be developed through family support programs. For instance, strengthening mutuality and parental concern within the family—such as through regular family activities or encouraging parents to express emotional support—can create a safer emotional environment for adolescents, thereby bolstering their psychological resilience [40]. Specifically, schools and communities could organize initiatives like “Parent-Child Co-Creation Days” or “Family Support Groups,” utilizing role-playing or collaborative family tasks (e.g., setting shared family goals) to foster mutual understanding and support among family members. Such family support programs not only reflect certain aspects of family functioning (e.g., emotional involvement and communication) but also provide positive outcomes for adolescents facing negative experiences and emotions [47]. Such strategies align with the dynamics of adaptive parent-child relationships and can effectively mitigate the impact of family dysfunction during adolescence [46]. These intervention strategies are associated with the mediating role of family functioning identified in this study, potentially offering effective pathways to support adolescent mental health.

2. Regarding the negative correlation between family functioning and emotional competence, communication training or conflict mediation may mitigate the associated impact of family dysfunction on emotional competence. For example, parents and adolescents could participate in communication training courses based on cognitive-behavioral approaches, learning to express emotional needs, listen to each other’s perspectives, and practice mediation skills in simulated conflict scenarios (e.g., using the “I-message” technique: “I feel angry because I think you’re not listening to me”). Such interventions may be associated with reduced family conflicts (e.g., issues in the conflict and harmony dimension), which may, in turn, be related to enhanced adolescent emotion regulation abilities, particularly when improving parent-child interactions through communication training. This is especially relevant for father-adolescent relationships, as Mastrotheodoros et al. [46] found that perceptual differences in conflict intensity between fathers and adolescents increase from mid- to late adolescence, suggesting that fathers may underestimate the emotional impact of conflicts on teens. Thus, interventions could include targeted modules for fathers, such as “Father Role Workshops,” which use case discussions and emotional education to help fathers better understand adolescents’ needs during puberty (e.g., autonomy), thereby reducing perceptual gaps and enhancing family functioning. Liu et al. noted that family support plays a moderating role in the relationship between bullying victimization and internet addiction among adolescents, suggesting that strengthening family support (e.g., through emotional communication and understanding) can buffer adolescents’ emotional distress in conflicts or negative experiences, thereby indirectly enhancing emotional competence [48].

3. Based on the 9.2% mediation effect identified in this study, comprehensive interventions aim to optimize family functioning and support psychological resilience and emotional competence, providing preliminary support for the mental health of students in grades 7–9. Comprehensive interventions should simultaneously focus on fostering psychological resilience and optimizing family functioning, particularly among students in grades 7–9 (approximately 13–15 years old), with the potential to be associated with enhanced emotional competence. This period marks a critical transition from early to mid-adolescence, where family dysfunction may intensify due to conflicts between adolescents’ growing need for autonomy and parental control [49]. Additionally, Mastrotheodoros et al. [46] noted that adolescents’ perception of conflict significantly increases during this stage, while parents—especially fathers—may remain unaware of these changes, leading to strained family relationships. Thus, school educators and counselors should collaborate with families to design targeted interventions. For instance, a “Psychological Resilience Bootcamp” could be introduced, incorporating team-building activities (e.g., outdoor challenges) and reflective exercises (e.g., maintaining a “resilience journal”) to help adolescents develop positive coping strategies and strengthen their resilience against family conflicts. Simultaneously, family intervention programs should focus on optimizing the five dimensions of family functioning (e.g., improving communication through family meetings and reducing excessive control through parental education), which may, in turn, be associated with enhanced emotional competence. Furthermore, family support may be related to improved adolescent psychological adaptation. Liu et al. found that family support enhances adolescents’ psychological adaptation in negative situations by moderating the relationship between anxiety and internet addiction [47]. This finding suggests that intervention programs should incorporate family support elements, such as parent-child communication workshops to strengthen emotional bonds, thereby optimizing family functioning and adolescents’ emotional competence. Additionally, considering that students in grades 7–9 are in a stage of rapid emotional development and self-awareness changes, interventions should adopt participatory and experiential methods. For example, through role-playing to simulate family scenarios, adolescents and parents can jointly explore ways to manage conflict and stress, potentially associated with enhanced psychological resilience and emotional competence. These intervention strategies are associated with the mediating role of family functioning identified in this study, potentially offering effective pathways to support adolescent mental health.

4. To ensure the sustainability of intervention effects, it is recommended to establish long-term tracking and evaluation mechanisms within intervention programs to validate their association with enhanced psychological resilience and emotional competence. For example, standardized scales (e.g., the Psychological Resilience Scale, Emotional Competence Scale, and Family Functioning Scale) could be administered before and after the intervention, as well as at 6-month and 12-month follow-ups, to assess effectiveness. Additionally, interviews or focus groups could be conducted to capture the subjective experiences of family members. Intervention designs should also account for differences in gender and family background (e.g., urban vs. rural). For rural families, which may face economic pressures or resource scarcity, supplementary community support—such as free psychological counseling services—could be provided. Liu et al.’s research indicates that family support is particularly crucial for rural adolescents, effectively mitigating the negative impact of bullying victimization on mental health. This suggests that intervention programs for rural families should prioritize resource support and the establishment of emotional bonds [48]. Through these comprehensive measures, interventions may be associated with enhanced psychological resilience and emotional competence among students in grades 7–9, while potentially supporting long-term improvements in family relationships, which may, in turn, be related to adolescents’ future social adaptation and mental health development.

Limitations and future directions

This study, using a large-sample cross-sectional design, has provided a preliminary understanding of the correlations among family functioning, psychological resilience, and emotional competence in primary and secondary school students in Chengdu, laying a foundation for future intervention practices and longitudinal research. However, the following limitations still exist:

First, regarding the research design, the cross-sectional approach cannot infer causal relationships between variables. The association pathways identified in this study may be bidirectional. For instance, not only might family dysfunction be related to lower emotional competence, but adolescents with poorer emotional competence might also perceive their family atmosphere more negatively, and their emotional problems could themselves exacerbate family conflicts. Furthermore, the possibility of unmeasured third variables (such as academic stress or peer relationships) being simultaneously correlated with these three core variables cannot be excluded. Future research could adopt a longitudinal tracking design to more clearly reveal the dynamic development and mutual influence mechanisms among family functioning, psychological resilience, and emotional competence.

Second, concerning the data source, this study relied primarily on student self-reports, which may be influenced by social desirability or recall bias. Although the Harman’s single-factor test indicated that common method bias was not severe, its potential influence cannot be entirely ruled out since all data were collected from student self-assessments at a single point in time. This might have exaggerated the strength of the correlations between variables to some extent. Future research could incorporate multi-source reports, such as from parents and teachers, to cross-validate the findings and enhance the objectivity and reliability of the data.

Furthermore, regarding model construction, to ensure an adequate fit of the Structural Equation Model (SEM) to the data, this study specified error covariances based on the overlap in item content. We acknowledge that while this approach has a theoretical basis and significantly improves model fit, it may sacrifice the model’s parsimony and generalizability. Additionally, this study found that the mediating effect of family functioning accounted for 9.2% of the total effect. Although statistically significant, this also indicates that the majority of the association between psychological resilience and emotional competence (over 90%) is either direct or achieved through other mediators not measured in this study (such as social support, coping styles, etc.). Therefore, family functioning is only a partial mechanism explaining the relationship between the two, and this conclusion should be interpreted with caution. Future research could attempt to use second-order factor models that include sub-dimensions or build chain mediation models incorporating multiple potential mediating variables to more comprehensively explore the complex interaction mechanisms among these variables.

Finally, in terms of theoretical depth, this study only tested a single mediation path and did not delve into the moderating roles of demographic variables such as gender, urban-rural location, or grade level. These factors might alter the strength or even the direction of the association pathways found in this study. Future research could construct more complex moderated mediation models to deepen the understanding of the developmental pathways of adolescent mental health.

Looking ahead, this research could be expanded in several ways. First, increasing the sample size to include more diverse family types (e.g., single-parent families or families with left-behind children) and comparing differences across regions (e.g., urban vs. rural) or grade levels (e.g., grades 7–9) would enhance the generalizability and representativeness of the findings. Second, incorporating variables such as adolescents’ personality traits or parents’ mental health status could explore their potential associations with family functioning, psychological resilience, and emotional competence. Furthermore, future research could focus on the characteristics of family functioning within the Chinese cultural context, such as the associations of filial piety or collectivism with parent-child relationships. Particularly in the context of Chinese collectivist culture, exploring the associations of filial piety and family harmony with adolescent psychological development could further enrich the theoretical framework. Through these improvements, a more comprehensive understanding of the mediating mechanisms of family functioning between psychological resilience and emotional competence can be achieved, offering more targeted evidence for adolescent mental health interventions. The preliminary findings of this study (9.2% mediation effect) provide exploratory guidance for adolescent mental health interventions, and future research could further validate and expand these findings through more complex designs.

Authors’ contributions

Xi Huang and Xiujuan Zhang spearheaded the study design, data collection, analysis, interpretation, and drafting of the manuscript. Qiong Chen and Yanling Hu assisted with the questionnaire survey. Xiufang Zhao and Li Zhao provided critical review and revisions to the manuscript. All authors are accountable for the content of the manuscript and have approved its final version for submission.

Funding

Hong Kong Polytechnic University and Sichuan University (19H0642) support.

this research.

Data availability

The datasets generated and/or analysed during the current study are available reasonable request to the corresponding author.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the guidelines of the Declaration of Helsinki and received approval from the Medical Ethics Committee of Sichuan University (Approval No. K2020025). Written informed consent was obtained from all participants and their legal guardians prior to their involvement in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.