Associations between parental marital conflict and short-video addiction among Chinese adolescents: a cross-sectional chain mediation analysis of experiential avoidance and anxiety

Abstract

This cross-sectional study examined the association between parental marital conflict and short-video addiction among adolescents and assessed indirect associations involving experiential avoidance and anxiety. The Children’s Perception of Interparental Conflict Scale, Acceptance and Action Questionnaire-II, Generalized Anxiety Disorder-2 Scale, and Short-Video Addiction Scale were administered to 4,065 adolescents from six schools across five provinces in China. PROCESS Model 6 and 5,000 bootstrap resamples were used to test the mediation effects. (1) Parental marital conflict, experiential avoidance, anxiety, and short-video addiction were significantly and positively correlated with one another (r = 0.280 – 0.553, p < 0.001). (2) The direct association between parental marital conflict and adolescent short-video addiction remained significant after the mediators were included in the model (β = 0.088, p < 0.001). (3) Experiential avoidance and anxiety independently mediated the association between parental marital conflict and short-video addiction and also formed a significant chain mediation pathway through experiential avoidance followed by anxiety. The total indirect effect accounted for 67.16% of the total effect. The results indicate that parental marital conflict was associated with short-video addiction both directly and indirectly through experiential avoidance, anxiety, and the prespecified chain pathway from experiential avoidance to anxiety. The findings broaden the evidence on family and emotional correlates of short-video addiction and highlight the relevance of supportive family relationships, emotion-regulation skills, media literacy, and adolescent psychological well-being.

Background

Short-video addiction (SVA) describes persistent and excessive engagement with short videos accompanied by diminished self-control, behavioral disruption, and attention-related difficulties. These problems can interfere with social functioning, academic performance, and occupational adaptation¹. Adolescents may be especially vulnerable. In a sample of 1,346 adolescent TikTok users, addictive use was accompanied by poorer mental health and school adjustment². A stratified Chinese survey reported severe SVA tendencies in approximately 14.8% of high school students³. Reported correlates of SVA include poorer sleep⁴, greater social anxiety⁵, increased academic procrastination⁶, and impaired cognitive functioning⁷. From an emotion-regulation perspective, adolescents may use short videos to distance themselves from internal conflict or distressing experiences, including depression and anxiety⁸. Accordingly, it is important to examine the pattern of associations linking family relational factors, emotional coping tendencies, and adolescents’ short-video use.

Research has increasingly linked adverse family environments with adolescents’ SVA, but the specific association with parental marital conflict requires closer attention. Parental marital conflict (PMC) encompasses verbal or physical disputes and aggression between spouses that arise from disagreements or related causes⁹. Many children are exposed to interparental conflict. In Australia alone, at least one million children are estimated to experience such conflict each year¹⁰. Evidence from urban households also indicates substantial exposure, with physical conflict reported in about 34.8% of families and psychological conflict in as many as 55.6%¹¹. As a recurring family stressor, PMC is relevant to adolescent emotional and behavioral adjustment. Social learning theory proposes that behavior is shaped not only by direct experience but also by observing and imitating others¹². In homes characterized by persistent tension and limited emotional support, adolescents may experience greater emotional distress and may turn to short videos as an accessible means of distraction and temporary emotional relief. The rapid feedback and reinforcement offered by these platforms may further encourage frequent and addictive use. A questionnaire study of 1,735 Chinese middle school students similarly found that greater PMC was associated with more severe SVA^13,14. Accordingly, the study proposed a positive association between PMC and adolescents’ SVA (H1).

When examining the psychological correlates of SVA, individual coping responses to emotional distress are especially relevant. Experiential avoidance (EA) denotes attempts to avoid or control situations, memories, thoughts, and feelings connected with unwanted internal experiences¹⁵. Emotional security theory holds that children seek to preserve a sense of safety when exposed to PMC. Threats to that security may encourage withdrawal, evasion, or other avoidant responses, which are consistent with EA¹⁶. PMC may therefore be positively associated with EA among adolescents. Prior evidence links EA with problematic internet, smartphone, and video-game use and identifies it as a correlate of behavioral addictions. The self-medication hypothesis further suggests that people may use particular behaviors to obtain temporary relief from internal distress¹⁷. Because short-video platforms provide intense stimulation and immediate feedback, adolescents with greater EA may rely on them to redirect attention from negative emotions. Empirical findings also associate EA with social-media use and smartphone addiction¹⁸. The study therefore proposed an indirect association between PMC and SVA through EA (H2).

Repeated avoidance can leave distressing emotional material insufficiently processed and may be accompanied by persistent anxiety. Anxiety is a pervasive and sustained negative affective state involving anticipatory worry, heightened vigilance, and physiological tension in response to uncertain future threats¹⁹. From 1990 to 2021, anxiety-disorder prevalence among people aged 10–24 years increased by 52%, with the largest rises among those aged 10–14 years and after 2019²⁰. A Chinese survey of 148,892 students aged 9–18 years reported high psychological distress in 31.6% of participants and approximately 38.8% of senior high school students²¹. The cognitive-behavioral model emphasizes that interpretations of events and coping responses are closely related to emotional outcomes. In the context of PMC, negative appraisals and greater threat sensitivity may be associated with continuing anxiety²². Thus, anxiety may constitute one of the psychological correlates linking parental marital conflict with adolescents’ SVA. Meta-analytic evidence also links stronger family conflict with greater adolescent anxiety²³. Anxiety is likewise associated with SVA. Uses and gratifications theory suggests that anxious adolescents may seek short videos for distraction and temporary emotional relief²⁴. Studies have further linked anxiety levels with the frequency and severity of short-video use^25,26. The study therefore proposed an indirect association between PMC and SVA through anxiety (H3).

PMC may also be related to how adolescents respond to distressing internal experiences. Emotional security theory suggests that threats arising from interparental conflict can encourage avoidance, suppression, or disengagement as attempts to obtain short-term relief¹⁶. Such responses do not remove the stressor and may limit effective emotional processing, emotion regulation, and problem solving. Continued reliance on EA may therefore coincide with tension, worry, and other anxiety responses. Prior research similarly identifies EA as a psychological vulnerability associated with anxiety symptoms²⁷. As anxiety rises, adolescents may increasingly use the immediate feedback, immersion, and attentional diversion of short videos to ease distress¹⁷. On this basis, the present study conceptualized experiential avoidance as a maladaptive coping correlate of parental marital conflict and anxiety as an emotional correlate that may accompany this coping pattern. Hypothesis 4 accordingly specified an indirect association between parental marital conflict and adolescents’ SVA through the ordered statistical path of experiential avoidance followed by anxiety. The ordering was defined for analysis and did not establish temporal priority or causality.

In summary, to clarify the pattern of associations among parental marital conflict, experiential avoidance, anxiety, and adolescents’ SVA, the present study tested a prespecified chain mediation model. Because both mediators were measured concurrently, a parallel mediation model and a reversed chain mediation model, with anxiety placed before experiential avoidance, were also tested to assess whether the indirect associations depended on the prespecified ordering.

Methods

Participants

This cross-sectional study was conducted between September and December 2024. A combination of convenience sampling and class-based cluster sampling was employed. First, six schools located in Hunan, Sichuan, Shandong, Yunnan, and Anhui Provinces were selected through convenience sampling, including two schools in Hunan Province and one school in each of the other four provinces. Subsequently, class-based cluster sampling was conducted within each participating school. Eligible participants were students enrolled in the selected classes who participated voluntarily, were able to complete the questionnaire independently, whose parent or legal guardian provided informed consent, and who provided adolescent assent. Before data collection, study information and electronic informed consent forms were distributed to parents or legal guardians through class-based parent communication groups. The study aims, procedures, voluntary nature of participation, anonymity, confidentiality, and right to withdraw at any time without penalty were also explained to the students before adolescent assent was obtained. Paper-and-pencil questionnaires were administered in classrooms under the supervision of class teachers. No monetary compensation was provided. A total of 4,245 students were initially recruited. Questionnaires were excluded if they contained unanswered items, showed identical responses across an entire scale or several consecutive scales, or exhibited obvious and highly repetitive response patterns. After 180 invalid questionnaires were excluded, 4,065 participants were retained, yielding a valid response rate of 95.76%. No missing values remained in the final analytical dataset; therefore, no missing-data imputation was performed. The final sample comprised 2,023 boys (49.77%) and 2,042 girls (50.23%), with a mean age of 14.71 years (SD = 1.43). Of these participants, 154 (3.79%) were primary-school students, 1,834 (45.12%) were junior-high-school students, and 2,077 (51.09%) were senior-high-school students. An a priori power analysis conducted using G*Power 3.1 indicated that a minimum sample of 153 participants was required, assuming an effect size of f² = 0.15, α = 0.05, statistical power of 0.95, and seven predictors. Thus, the final sample substantially exceeded the minimum required sample size.

Measures

Parental marital conflict

PMC was assessed with the Children’s Perception of Interparental Conflict Scale (CPIC)²⁸, using the version adapted for Chinese participants²⁹. The five items are scored on a four-point Likert scale from 1 (strongly disagree) to 4 (strongly agree), producing a total score of 5–20. Higher totals indicate greater conflict as perceived by the child. In the present sample, internal consistency was acceptable (Cronbach’s α = 0.776).

Experiential avoidance

Experiential avoidance was assessed using the Chinese version of the Acceptance and Action Questionnaire-II (AAQ-II). The original AAQ-II was developed by Bond et al.³⁰ as a revised version of the AAQ-I and was subsequently translated and validated among Chinese university students by Cao et al.³¹. The scale consists of seven items rated on a 7-point Likert scale ranging from 1 (“never true”) to 7 (“always true”). The item scores are summed to produce a total score ranging from 7 to 49, with higher scores indicating greater experiential avoidance or psychological inflexibility. Cao et al. reported good internal consistency for the Chinese version (Cronbach’s α = 0.88), and confirmatory factor analysis supported an acceptable one-factor structure. In the present sample, the AAQ-II demonstrated good internal consistency, with a Cronbach’s α of 0.916.

Anxiety

Anxiety symptoms were measured with the two-item Generalized Anxiety Disorder-2 scale (GAD-2)³². Each item has four response options. Standard GAD-2 scoring ranges from 0 (not at all) to 3 (nearly every day); for data-entry convenience, this questionnaire coded the same options from 1 to 4. This shift adds one point to each item, yielding a recorded total of 2–8. Subtracting 2 converts the recorded value to the standard 0–6 range. Higher scores denote more severe anxiety symptoms. Cronbach’s α was 0.852 in the present sample.

Short-video addiction

Short-video addiction was assessed with the SVA Scale developed by Mao et al.³³ and validated in adolescent samples³⁴. Its 13 items cover 3 dimensions: cognitive-behavioral changes, physical impairment, and social engagement. Responses are given on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree), and higher totals indicate more severe addictive tendencies. Internal consistency in this sample was good (Cronbach’s α = 0.870).

Statistical analysis

Statistical analyses were conducted using SPSS 27.0. Descriptive statistics, Pearson correlations, and Harman’s single-factor test were performed. Confirmatory factor analysis was conducted using Amos 29.0 with maximum likelihood estimation to compare the hypothesized six-factor model with a single-factor model.

Before the mediation analysis, all continuous variables were standardized. PROCESS Model 6 was used to test the prespecified chain mediation model involving experiential avoidance and anxiety in the association between PMC and SVA³⁵. All regression models were adjusted for gender, age, and school stage. Gender was entered as a binary variable, age as a continuous variable, and school stage was represented by two dummy variables corresponding to primary school, junior high school, and senior high school. All indirect effects and their 95% confidence intervals were estimated using 5,000 bootstrap resamples³⁶. An indirect effect was considered statistically significant when the corresponding 95% bootstrap confidence interval did not include zero. All statistical tests were two-sided, with p < 0.05 indicating statistical significance.

Because experiential avoidance and anxiety were measured at the same time point and other theoretically plausible mediator specifications could not be ruled out, two alternative-model robustness analyses were additionally conducted. First, PROCESS Model 4 was used to estimate a parallel mediation model in which experiential avoidance and anxiety were simultaneously specified as mediators without a directional path between them. Second, PROCESS Model 6 was used to estimate a reversed chain mediation model in which anxiety was specified as the first mediator and experiential avoidance as the second mediator. Both alternative models used the same analytical sample, covariates, variable-processing procedures, and 5,000 bootstrap resamples. These analyses examined whether the observed indirect associations depended on the prespecified ordering of the mediators rather than establishing temporal precedence or causal relationships.

Results

Common method bias test

Potential common method bias was evaluated because all variables came from self-report questionnaires. Harman’s single-factor test was supplemented with confirmatory factor analysis comparing the hypothesized six-factor structure with a single-factor alternative. The unrotated exploratory analysis extracted multiple factors with eigenvalues above 1; the first explained 33.61% of the variance, below the 40% criterion. The six-factor model represented PMC, experiential avoidance, anxiety, and the three SVA dimensions—cognitive-behavioral changes, physical impairment, and social engagement—as correlated latent factors, with each item assigned to its intended construct. The alternative model loaded every item onto one latent factor. Fit was poor for the single-factor model (χ²/df = 75.97, CFI = 0.551, TLI = 0.514, RMSEA = 0.136, SRMR = 0.135) and substantially better for the six-factor model (χ²/df = 11.85, CFI = 0.938, TLI = 0.930, RMSEA = 0.052, SRMR = 0.035). Taken together, these results indicate that one common factor did not adequately represent the measurement items and that common method bias was unlikely to materially distort the findings.

Descriptive analysis

Table 1 shows that girls had higher mean scores than boys for PMC, experiential avoidance, anxiety, and SVA. All four variables also differed significantly across school stages.

Table 1.

Describes the analysis.

Variables		PMC			EA		Anxiety		SVA
		Mean		SD	Mean	SD	Mean	SD	Mean	SD
Gender	Boys		10.25	3.53	20.89	10.43	3.69	1.67	35.89	10.96
	Girls		10.68	3.46	22.33	10.27	3.96	1.68	37.80	10.74
	t		−3.93***		−4.44***		−5.14***		−5.60***
Grade	Primary school		10.32	3.53	21.45	9.90	3.90	1.79	37.07	11.04
	Junior high school		10.14	3.54	20.24	10.47	3.68	1.70	35.44	11.04
	Senior high school		10.76	3.44	22.85	10.17	3.94	1.65	38.08	10.59
	F		15.60***		31.33***		11.35***		29.22***

***: p<0.001.

Correlation analysis

Table 2 indicates positive correlations between PMC and experiential avoidance, anxiety, and SVA. Experiential avoidance was positively correlated with anxiety and SVA, and anxiety was positively correlated with SVA.

Table 2.

Correlation analysis.

Variables	1	2	3	4
1 PMC	-
2 EA	0.525***	-
3 Anxiety	0.340***	0.553***	-
4 SVA	0.280***	0.397***	0.313***	-

Note.***: p<0.001.

Mediation model test

With demographic variables controlled, Table 3and Fig. 1 indicate a positive association between interparental conflict and SVA. The direct association remained significant after experiential avoidance and anxiety were entered as mediators. Interparental conflict was also positively associated with both mediators. Experiential avoidance was positively associated with anxiety and SVA, while anxiety was positively associated with SVA. Table 4 presents the contribution of each indirect pathway to the total effect.

Table 3.

Regression results for the mediation model.

Outcome variables	Explanatory variables	β	SE	t	R²	F
SVA	PMC	0.268	0.015	17.836***	0.096	107.67***
EA	PMC	0.516	0.013	38.709***	0.286	406.015***
Anxiety	PMC	0.066	0.015	4.315***	0.311	366.931***
	EA	0.518	0.015	33.583***	0.311	366.931***
SVA	PMC	0.088	0.017	5.255***	0.186	154.314***
	EA	0.266	0.019	14.063***	0.186	154.314***
	Anxiety	0.127	0.017	7.433***	0.186	154.314***

Note. All continuous variables were standardized before the regression analyses. β represents the standardized regression coefficient. PMC = parental marital conflict; EA = experiential avoidance; SVA = short-video addiction. ***p < 0.001.

Fig. 1.

The chained mediation model.

Table 4.

Decomposition of direct and indirect associations.

Intermediate path	Standardized effect	SE	Bootstrap 95% CI	Proportion of mediating effect
Total effect	0.268	0.015	0.238, 0.297
Direct effect	0.088	0.017	0.055, 0.121
Total indirect effect	0.180	0.011	0.158, 0.203	67.16%
PMC→EA →SVA	0.138	0.012	0.115, 0.161	51.49%
PMC → Anxiety→ SVA	0.008	0.002	0.004, 0.014	2.99%
PMC → EA→ Anxiety → SVA	0.034	0.005	0.024, 0.045	12.69%

Note. CI = confidence interval; PMC = parental marital conflict; EA = experiential avoidance; SVA = short-video addiction.

Alternative model testing

To examine whether the indirect associations depended on the prespecified ordering of experiential avoidance and anxiety, we further tested a parallel mediation model and a reversed chain mediation model. In the parallel mediation model, the indirect effects through experiential avoidance and anxiety were both significant. In the reversed chain mediation model, the pathway from parental marital conflict to SVA through anxiety and then experiential avoidance was also significant (Table 5).

Table 5.

Results of alternative model testing.

Model	Indirect pathway	Effect	Bootstrap SE	95% bootstrap CI
Parallel mediation model	PMC → EA → SVA	0.137	0.012	[0.114, 0.160]
	PMC → Anxiety → SVA	0.042	0.007	[0.030, 0.055]
	Total indirect effect	0.179	0.011	[0.157, 0.202]
Reversed chain mediation model	PMC → Anxiety → SVA	0.042	0.007	[0.030, 0.055]
	PMC → EA → SVA	0.100	0.009	[0.083, 0.117]
	PMC → Anxiety → EA → SVA	0.037	0.004	[0.030, 0.045]
	Total indirect effect	0.179	0.011	[0.157, 0.202]

Note. All alternative models used the same sample, covariates, standardized continuous variables, and 5,000 bootstrap resamples as the hypothesized model. PMC = parental marital conflict; EA = experiential avoidance; SVA = short-video addiction.

Discussion

The present cross-sectional analysis examined associations among PMC, EA, anxiety, and SVA and evaluated the indirect associations involving EA and anxiety. PMC was positively correlated with SVA and with both mediators. Significant specific indirect associations were observed through experiential avoidance and anxiety, together with the prespecified chain path from experiential avoidance to anxiety. The results were consistent with all proposed hypotheses.

Previous studies have repeatedly documented a positive association between PMC and SVA. Adolescents in conflictual family settings often experience greater emotional distress, which is also related to heavier reliance on short videos³⁷. Recent evidence likewise connects PMC with emotional difficulties and problematic digital-media use¹⁴. Adolescents living in high-conflict homes have also shown higher levels of SVA³⁸. The present result therefore supports Hypothesis 1.

PMC was also positively correlated with EA, consistent with earlier work³⁹. Ongoing exposure to family conflict may be accompanied by greater use of avoidance-oriented coping to reduce distress or disengage from upsetting situations⁴⁰. EA, in turn, was positively correlated with SVA, in line with prior findings⁴¹. Habitual efforts to suppress or avoid internal experiences may increase reliance on external stimulation for distraction. Short-video platforms are especially suited to this pattern because they provide immediate feedback and strong sensory input⁴². The significant indirect association through experiential avoidance supports Hypothesis 2.

Anxiety also showed a significant indirect association between parental marital conflict and adolescents’ short-video addiction, supporting Hypothesis 3. Prior research has consistently associated parental marital conflict with adolescent anxiety²³, and the present findings showed the same positive pattern. From the perspectives of the cognitive-behavioral model and uses and gratifications theory, PMC may be related to greater threat sensitivity, tension, and worry. Adolescents experiencing more anxiety may use short videos to redirect attention and obtain temporary relief. Longitudinal evidence has linked anxiety symptoms with later increases in short-video addiction²⁶, while individuals with addictive short-video use have also reported higher anxiety². Anxiety therefore appears to be an important emotional correlate connecting parental marital conflict with short-video addiction.

The prespecified chain indirect association from PMC to SVA through experiential avoidance and then anxiety was significant, supporting Hypothesis 4. Experiential avoidance was also positively associated with anxiety, consistent with earlier evidence⁴³. EA reflects attempts to avoid distressing internal experiences in order to obtain short-term relief. Although this response may feel helpful initially, it can restrict effective emotion regulation and increase internal conflict and psychological burden⁴⁴. Greater reliance on experiential avoidance has been associated with persistent tension, worry, and higher levels of anxiety⁴⁵. Relevant empirical studies have shown that avoidance strategies do not eliminate underlying stressors and are associated with accumulated emotional distress and more severe anxiety and depressive symptoms⁴⁶. Alternative-model analyses also yielded significant indirect associations in the parallel and reversed chain mediation specifications. These findings indicate that the cross-sectional data were compatible with more than one mediator specification and did not uniquely identify the ordering of experiential avoidance and anxiety. Although the hypothesized pathway from experiential avoidance to anxiety is theoretically plausible, the cross-sectional design precludes conclusions regarding temporal precedence. Accordingly, the proposed mediator ordering should be interpreted as a theoretically informed representation of the observed associations rather than evidence of a causal sequence.

Evidence from different cultural contexts is broadly consistent with the present findings. A study of Italian adolescents found that interparental conflict was associated with emotional and behavioral difficulties, with psychological inflexibility showing a significant indirect association between these variables³⁹. Research conducted in Japan further indicated that adolescents’ coping responses to interparental conflict and emotional insecurity were closely associated with their psychological adjustment⁴⁷. A Spanish study found that experiential avoidance was positively associated with problematic Internet, mobile-phone, and video-game use among adolescents⁵. Similarly, research involving Greek adolescents showed that problematic TikTok use was associated with higher levels of anxiety, depression, and sleepiness⁴⁸. These findings suggest that interparental conflict, avoidance-related coping, and negative emotional states are relevant correlates of adolescents’ psychological and digital-media-related problems across different cultural contexts. The present study extends this evidence by examining PMC, experiential avoidance, anxiety, and SVA within a single statistical model among Chinese adolescents. However, the strength and specific patterns of these associations may vary across sociocultural contexts, and further cross-cultural research using comparable measures is warranted.

The combined analysis of parental marital conflict, experiential avoidance, anxiety, and short-video addiction revealed a coherent pattern of direct and indirect associations. Prevention initiatives involving families, schools, and mental-health services may therefore consider family conflict, experiential avoidance, anxiety, emotion regulation, and media literacy together. Because the study was cross-sectional, longitudinal and intervention research is required before concluding that changes in these factors will be followed by lower short-video addiction. Beyond family-related factors, religiosity and spirituality may constitute additional contextual influences on adolescents’ emotional adjustment and digital-media behaviors. A recent systematic review indicated that religiosity was frequently associated with a lower risk of problematic or addictive technology use, although religion-related experiences may also be linked to psychological distress when adolescents encounter identity-based exclusion, harassment, or conflict. These findings suggest that the roles of religiosity and spirituality may vary according to their specific dimensions and sociocultural contexts. Because these factors were not assessed in the present study, future research should examine whether they exert independent or moderating effects on the associations among parental marital conflict, experiential avoidance, anxiety, and SVA⁴⁹.

Placing PMC, experiential avoidance, anxiety, and SVA in one statistical model offers an integrated view of family and emotional correlates of problematic short-video use. Several limitations should be considered. First, all study variables were assessed concurrently. Consequently, the temporal ordering among parental marital conflict, experiential avoidance, anxiety, and short-video addiction could not be determined. PROCESS Model 6 estimates indirect associations based on a prespecified statistical ordering of variables, but it does not demonstrate that experiential avoidance temporally preceded anxiety or that the observed associations represented causal mechanisms. The significant parallel and reversed chain mediation models further indicate that alternative specifications are also compatible with the data. Future longitudinal, cross-lagged, or experimental studies are needed to examine temporal ordering and evaluate possible causal processes. Second, reliance on adolescent self-report may have introduced recall error, social-desirability effects, and shared-method variance. Future work should combine adolescent responses with parent or teacher reports, clinical assessments, and objective records of digital behavior. Third, adolescent data are inherently hierarchical, as students are nested within families, classes, and schools. Individuals within the same family or school may share similar socioeconomic conditions, family environments, and educational experiences; therefore, their observations may not be fully independent. Multilevel analyses are important because they allow individual-, family-, and school-level sources of variation to be distinguished and facilitate a more comprehensive examination of parental and family factors associated with adolescent psychological and behavioral outcomes⁵⁰. However, the present study included only six school-level clusters and did not collect family-level identifiers, which limited the reliable estimation of school- and family-level random effects. Future studies should recruit larger numbers of schools and families and employ multilevel models or appropriately adjusted cluster-robust methods. Fourth, the schools were selected through convenience sampling, and the sample was not nationally representative. Although participants were recruited from five provinces, the generalizability of the findings to adolescents from other regions and different sociocultural contexts remains limited. Future studies should use probability-based sampling and recruit more geographically diverse samples to improve the representativeness and external validity of the findings. Finally, several potentially important confounding variables, including socioeconomic status, current parental marital status, depressive symptoms, and the duration or frequency of short-video use, were not assessed. Their omission may have resulted in residual confounding. Future studies should measure and adjust for these factors.

Conclusion

PMC was positively associated with adolescents’ SVA, while experiential avoidance and anxiety showed significant independent indirect associations and a significant chain indirect association. Support for the alternative models indicates that the relation between experiential avoidance and anxiety cannot be treated as exclusively one-directional in these cross-sectional data. The findings add to evidence on family and emotional correlates of SVA and suggest that prevention should extend beyond limits on screen use. Coordinated efforts may focus on reducing adolescents’ exposure to parental conflict, strengthening psychological flexibility and emotion-regulation skills, recognizing anxiety symptoms, and improving media literacy across families, schools, and mental-health services.

Author contributions

Yuhan Chen1245, Xiaodi Ma1256, Gen Li1356, Yang Liu123456.1 Conceptualization; 2 Methodology; 3 Data curation; 4 Writing - Original Draft; 5 Writing - Review & Editing; 6 Funding acquisition.

Funding

2025 Graduate Research Innovation Project of Jishou University: JDY2025043.

Data availability

The datasets generated and/or analysed during the current study are not publicly available due [our experimental team’s policy] but are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

The study was approved by the Biomedicine Ethics Committee of Jishou University before the initiation of the project (JSDX-2024-0086). Informed consent was obtained from the participants before the start of the program. The study was conducted in accordance with the Declaration of Helsinki.