An initial sample of 1,100 junior middle school students in grades 7 and 9 were recruited from three junior middle schools in Beijing, Chongqing, and Shijiazhuang, three cities in China, 62 of whom were excluded from the data analysis due to incomplete data (attrition rate = 5.64%). The final sample consisted of 1,038 students (480 boys and 554 girls, and four gender missing). The proportions of each grade were 46.1, 41.6, and 12.2%, respectively. All participation was voluntary. Prior to data collection, the participating schools, the student participants and their parents were informed about the research’s objectives, confidentiality and anonymity. According to the information provided by the participating students, 96.43% of them have intact families, i.e., their parents are not separated or divorced. During data collection, assessments of the study variables (i.e., family functioning, parental conflict, shyness, and loneliness) were administered to the participants using standard instructions in a group setting with ~40–50 participants in regular classrooms. The assessment procedure took ~50 min to complete.

SPSS 23.0 was used for data entry and management; SPSS 23.0 and AMOS 17.0 were used for data analysis.

In terms of parental conflict, the perceived level of parental conflict among adolescents was generally at a low level (M_{parental conflict} = 1.03), and the difference between genders was not significant, but the difference between grades was significant. Specifically, the scores of the second-grade students were significantly higher than those of the first-grade students, while the differences between the second grade and the third grade or between the third grade and the first grade were not significant [F_{(2, 1,037)} = 3.31, p < 0.05; M_{second grade}-M_{first grade} = 0.67, p < 0.05; M_{second grade}-M_{third grade} = 0.12, p > 0.05; M_{third grade}-M_{first grade} = 0.54, p > 0.05].

In terms of socially adverse emotions, the shyness and loneliness of adolescents were generally at the lower-middle level (M_shyness = 2.61, M_loneliness = 3.93), and there were significant gender and grade differences. Specifically, boys were more shy than girls (M_boys = 34.83, M_girls = 33.18, t = 3.16, p < 0.01) and felt more lonely (M_boys = 8.14, M_girls = 7.60, t = 2.01, p < 0.05). In terms of shyness, there was no significant grade difference [F_{(2, 1,038)} = 1.25, p > 0.05]. In terms of loneliness, compared with the first grade, the second grade experience more loneliness, while the difference between the first grade and the third grade, the second grade and the third grade was not significant [F_{(2, 1,038)} = 4.05, p < 0.05; M_{second grade}-M_{first grade} = 0.80, p < 0.05; M_{third grade}-M_{first grade} = 0.54, p > 0.05; M_{second grade}-M_{third grade} = 0.27, p > 0.05].

In terms of family functioning, the family closeness felt by teenagers was generally at the upper-middle level (M_Closeness = 4.31), and the difference between grades was not significant, but the difference between genders was significant. Specifically, girls felt significantly higher family closeness than boys (M_boys = 68.03, M_girls = 69.85, t = −2.46, p < 0.01). Teenagers felt family adaptability in general at the upper-middle level (M_adaptability = 3.39), and there were no significant differences between gender and grade. Table 1 shows the detailed results.

The results of the correlational analysis (see Table 2) showed that each dimension of parental conflict was significantly negatively correlated with each dimension of family functioning and significantly positively correlated with adolescents’ shyness and loneliness. This indicates that the more frequent and intense the conflict, the poorer the degree of conflict resolution, the lower the family cohesion and adaptability, and the higher the likelihood that teenagers will experience socially adverse emotions. At the same time, family closeness and adaptability were significantly negatively correlated with adolescents’ shyness and loneliness, indicating that the better family functioning is, the less likely teenagers are to experience socially adverse emotions.

First, the predictive effect of parental conflict on adolescents’ socially adverse emotions was detected. The results showed that χ²/df = 10.475, RMSEA = 0.096, NFI = 0.966, TLI = 0.939, CFI = 0.969, indicating that all the fitting indicators of the model reached relatively good levels. Based on the model results (see Figure 2), parental conflict had significant positive predictive effects on adolescents’ shyness (β = 0.200, p < 0.001) and loneliness experience (β = 0.212, p < 0.001).

Next, a structural equation model was created by using family functioning as a mediating variable to verify the fitting degree of the hypothesized mediation model. The results showed that χ²/df = 7.380, RMSEA = 0.078, GFI = 0.978, NFI = 0.971, and CFI = 0.975 all reached satisfactory levels (see Figure 3).

As shown in Figure 3, after family functioning, the mediating variable was added, the path coefficient from parental conflict to adolescents’ shyness was still significant (β = 0.150, p < 0.001), and parental conflict significantly negatively predicted family functioning (β = −503, p < 0.001), family functioning significantly negatively predicted adolescents’ shyness (β = −0.092, p < 0.05); when the mediating variable of family functioning was added, the path coefficient between parental conflict and adolescents’ loneliness was still significant (β = 0.075, p < 0.05), parental conflict significantly negatively predicted family functioning (β = −0.503, p < 0.001), and family functioning significantly negatively predicted adolescent’ loneliness (β = −0.272, p < 0.001).

The mediating effect test was performed using the bootstrap method with a 95% confidence interval. The sampling was repeated 5,000 times. The mediation effect was significant if the confidence interval did not contain 0. This study reached two significant mediation chains: parental conflict → family functioning → adolescents’ shyness and parental conflict → family functioning → adolescents’ loneliness. The mediating effect values were 0.05 and 0.14, respectively (see Table 3). Since the direct effect of the first mediation chain is still significant, the mediation is regarded as partial mediation. Since the direct effect of the second mediation chain is not significant, therefore, the mediation is complete. The results indicated that family functioning played a partial mediating role in the relationship between parental conflict and adolescents’ shyness and a complete mediating role in the relationship between parental conflict and adolescents’ loneliness.

In this study, the perceived parental conflict, the experience of shyness, and the experience of loneliness in the second grade in junior high school were higher than those in the first and third grades. This is consistent with the views of “problematic second grade in junior high school” and “key second grade in junior high school.” The particularity of the second grade in junior high school can be explained by the teenagers themselves and their parents.

From the perspective of the developmental characteristics of teenagers, those in the second grade of junior high school, who are 13 or 14 years old, are in the “storm period” of physical and mental development. At the level of psychological development, the pursuit of autonomy was the most important theme in this period. The pursuit of autonomy means the rapid development of independence and self-awareness of teenagers. On the one hand, this is manifested in rebellion at the behavioral level and the attempt to show their individuality; on the other hand, it is manifested in excessive self-focus at the conscious level, caring very much about others’ evaluations of themselves, and even many sometimes exaggerating the degree of attention others pay to them. This rebellion and self-focus, which may reach its peak in the second grade of junior high school, will cause friction or conflict with the outside world, causing teenagers to experience more loneliness. At the same time, individuals in the rebellious or self-focused state are prone to be irritated by little things and are prone to mistake subtle frictions for conflicts. Therefore, conflicts between their parents will increase to a higher level than the conflicts’ actual level.

In terms of physiological changes, male and female middle school students who entered adolescence began to show physiological changes, with the biggest changes among 13- and 14-year-old junior high school students. As sexual development begins to mature, teenagers gradually close themselves to the outside world. At the same time, their emotions toward the opposite sex gradually develop along with their physical and psychological development. They begin to have a hazy affection for the opposite sex. Focusing on the impression one has in the eyes of friends and the image among classmates of the opposite sex may also be an important reason why the second year of junior high school became a turning point for adolescents’ shy behavior and shy cognition.

From the perspective of parents, because the first grade of junior high school has just been promoted from elementary school to junior high school, the new environment and the new peer relationship has not yet been established, there are many problems in life, study, and psychological adaptation that require parents’ help and support. The third grade of junior high school is the top priority for many Chinese families. At these critical moments, to allow their children to adapt to junior high school life or prepare for the high school entrance examination with peace of mind, parents will consciously reduce the frequency of conflicts between them or reduce the intensity of conflicts if they cannot be avoided. In contrast, the second grade of junior high school is in the middle stage, and students have relatively fewer practical difficulties and pressures. Therefore, parents may not deliberately avoid conflicts, so teenagers in the second grade of junior high school perceive more parental conflict. According to the individual-environment interaction model, internalizing issues such as individual emotions will be affected by factors such as the family environment. The more frequent and severe the children’s perceived parental conflict, the more unfavorable the child’s healthy emotional development, which explained why the levels of shyness and loneliness in the second grade were higher than those in the first and third grades.

In terms of family functioning, girls reported significantly higher family cohesion than boys; in terms of socially adverse emotions, their experiences of shyness and loneliness were significantly lower than those of boys. Traditionally, girls were sensitive and fragile, while boys were dominant and competitive. Under the influence of this traditional concept, parents generally give girls more help and attention in life and give boys more restrictions and control, which to a certain extent makes girls perceive more closeness between family members than boys. Second, during adolescence, influenced by gender stereotypes, boys tend to take a tough stance and show their aggression in their dealings with others, while girls are encouraged to be friendly, cooperative, and value their relationships with others. Therefore, girls pay more attention to close relationships with family and show more pro-social behaviors, while boys have difficulty actively expressing their inner emotional distress and experience more shyness and loneliness.

From the results of the correlation analysis, it can be seen that the higher the level of parental conflict (i.e., the higher the intensity, the higher the frequency, and the poorer the conflict resolution), the higher the adolescents’ experience of shyness and loneliness. This validated the hypothesis of the present study and was consistent with previous studies (Hastings et al., 2015). Cognitive-situation theory can well explain the relationship between parental conflict and adolescents’ shyness. According to this theory, parental conflict makes adolescents feel threatened and even self-blame (Jordan, 2016). In the face of parental conflict, children tend to blame themselves for the discord between their parents, believing that the parent’s conflict is caused by themselves, and their self-evaluation will be reduced (He et al., 2023). Low self-evaluation can lead to children’s fear of others’ negative evaluation of them in the social process, and this is precisely the reason why children show shyness in social interactions. From the definition of loneliness, the reason why children feel lonely and are not accepted by their peers is because their behavioral problems such as aggression and withdrawal, as well as a lack of social interaction ability. According to Bandura’s social learning theory, parental conflict occurs either through the cold war or through quarrels, physical friction, or the way in which one party quarrels and the other party avoids or accommodates. Children who grow up in this environment for a long time will learn this kind of behavior and will be rejected by their peers and feel isolated and lonely.

Additionally, in the present study, the higher the level of parental conflict was, the lower the adolescents’ reported levels of family cohesion and adaptability. This was generally consistent with previous studies (Deng and Li, 2024). Family is a system, with the relationship between husband and wife as an important part of it. Frequent or high-intensity conflicts will not only destroy the emotional bond between two people but also break the harmonious atmosphere of the whole family and indirectly destroy the emotional bond between family members. In turn, the closeness between family members will be affected, and the family’s resilience will be reduced. In addition, many empirical studies have also confirmed that in families with high levels of parental conflict, children are more likely to be involved in a parent–child triangle relationship. In these relationships, children replace their parents in taking care of the family, while parents are more likely to beat and scold their children, forcing the children to become the “scapegoat” in the conflict between husband and wife. It is more likely that these kids will form an alliance with one parent against the other (Harman et al., 2022). These relationships confuse family roles and task division among family members, weaken the ability of the family to solve problems, and cause family functioning to begin to malfunction.

The results of structural equation modeling showed that the relationship between parental conflict and adolescents’ social impairment was mediated by family functioning, which was consistent with the research hypothesis and well supported the proximal-distal theory (Kroencke et al., 2023; Lionetti et al., 2024). An individual’s psychological development is the result of the continuous interaction between them and their surrounding environment; the distal factors in the ecosystem will affect the development of the individual through the proximal environment. Compared with parental conflict, which only reflects the relationship between parents, for adolescents, family functioning is the proximal environment that affects adolescents’ psychological development.

The results of the bootstrap analysis showed that family functioning played different mediating roles between parental conflict and adolescents’ shyness and loneliness. On the one hand, family functioning completely mediated the relationship between parental conflict and adolescents’ loneliness, i.e., after the family functioning factor was added, the direct effect of parental conflict on adolescents’ loneliness was no longer significant; on the other hand, parents’ family functioning played a partial mediating role in the relationship between conflict and adolescents’ shyness. That is, after the family functioning factor was included, the direct effect of parental conflict on adolescents’ shyness was still significant. As two important indicators of social distress in teenagers, why do family functioning have different effects on loneliness and shyness? This may be because compared to shyness, loneliness is more closely related to family functioning. Family and peers are two very important environmental factors that affect an individual’s loneliness. During adolescence, children have not yet been separated from their parents and the family is their main living environment, so the influence of family functioning, especially family closeness, on their loneliness becomes more important.

According to previous studies, there are many reasons for shyness. Parents’ neglect and indifference, individual poor attribution style and coping style, and unreasonable self-protection mechanisms are all important predictors of adolescents’ shyness. Family factors are only some of the factors. The results of the correlation analysis in this study also showed that the correlation coefficient between loneliness and each dimension of family functioning was significantly higher than that between shyness and each dimension of family functioning [t_(closeness) = 3.808, p < 0.001; t_{(adaptability)} = 3.499, p < 0.001]. However, in general, good family functioning has a positive effect on adolescents’ socially adverse emotions.