Abstract
Background
Against the backdrop of rising family educational expenditures and increasing prevalence of adolescent depression, this study examined the parallel mediating effects of psychological and material pathways on adolescent well-being.
Methods
This study analyzed data from 1151 adolescents aged 10–19 years in the 2022 China Family Panel Studies (CFPS). A parallel multiple mediation model was used to examine two pathways: a psychological pathway in which future orientation is associated with well-being via depressive symptoms, and a material pathway in which educational investment is associated with well-being via depressive symptoms. Indirect effects and differences between pathways were tested using bias-corrected bootstrapping with 5000 resamples.
Results
Future orientation was significantly negatively associated with depressive symptoms (β = –0.257, p < 0.001), with a significant indirect effect on well-being via depressive symptoms (indirect effect = 0.167, 95% CI [0.107, 0.238]). Family educational investment was not significantly associated with depressive symptoms (β = –0.036, p = 0.203), and its indirect effect was non-significant. The difference between the two indirect effects was significant (Δ = 0.157, 95% CI [0.101, 0.231]), indicating that the indirect effect via the psychological pathway was stronger than that via the material pathway.
Conclusion
These findings provide preliminary evidence that future orientation, as a psychological resource, is associated with lower depressive symptoms and higher hedonic well-being among Chinese adolescents. However, owing to the cross-sectional design and measurement limitations (eg, a single-item measure of eudaimonic well-being), the results should be interpreted as statistical associations rather than causal pathways. They do not justify broad claims that challenge human capital theory or provide strong support for hope theory; rather, they suggest that psychological and material pathways may differ in their associations with adolescent mental health. Future longitudinal and, where feasible, experimental or intervention research using independent measures is needed.
Keywords: future orientation, educational investment, depressive symptoms, well-being, parallel multiple mediation, adolescents
Introduction
Adolescent mental health has emerged as a global public health priority. The United Nations Children’s Fund (UNICEF) 2021 report indicated that approximately 14% of adolescents aged 10–19 years worldwide are experiencing various mental disorders, with depression and anxiety being the most prevalent psychological problems.1 This issue is equally concerning in China, where the detection rate of depression among adolescents is 24.6%, with severe depression accounting for 7.4%.2 Adolescent mental health not only affects individual development but also profoundly influences the accumulation of healthy human capital and the process of high-quality population development.3
Currently, family educational investment in China is continuously rising, with both monetary and non-monetary resources allocated to children’s development showing significant growth.4,5 From the perspective of human capital theory, family educational investment is conceptualized as a rational investment in enhancing children’s cognitive abilities and future developmental potential.6 However, this theoretical framework primarily focuses on cognitive and economic outcomes, with no direct consideration of mental health. In reality, increased educational investment has not been accompanied by synchronous improvements in adolescent mental health. In some cases, excessive family educational resource investment may instead increase adolescents’ psychological burden.7–10 This paradox reveals a fundamental dilemma in contemporary family educational investment: families are driven by cultural traditions and educational competition to invest continuously, yet they worry that excessive investment may harm their children’s mental health.11
Future orientation is defined as an individual’s expectations, plans, and subjective confidence regarding their own future developmental possibilities.12 As a key component of positive psychological resources, future orientation plays a crucial role in adolescent stress coping and mental health maintenance. Hope theory conceptualizes hope as a positive motivational state based on a sense of successful goal-directed determination, comprising two core dimensions: pathways thinking (perceiving routes to desired goals) and agency thinking (motivation to pursue those routes).13 Individuals with high future orientation can reframe current difficulties from a positive perspective, transforming stress into opportunities for growth.14 A longitudinal study of Swiss adolescents by Schmid et al15 found that positive future expectations were significantly associated with lower levels of depressive symptoms and higher self-regulation abilities.16,17 During adolescence in particular, hope and negative emotions have bidirectional relationships: hope reduces negative emotions, but negative emotions also erode hope levels.18,19
Adolescents globally are experiencing a profound reconstruction of future orientation.20 The Viable Youth Futures framework proposed by Dyson et al21 identifies dimensions of adolescent future thinking including survival concerns, cross-domain action, ethical reflection, and alternative imagination. However, these qualitative findings have yet to be tested in large-sample quantitative studies. More importantly, existing research rarely compares psychological resources (future orientation) and material resources (educational investment) within the same parallel mediation framework to assess their relative contributions.
To address this gap, the present study utilized data from the 2022 China Family Panel Studies (CFPS) to construct a parallel multiple mediation model systematically testing the relative explanatory power of psychological and material pathways to adolescent well-being. The findings provide preliminary empirical evidence and theoretical insights for adolescent mental health research and well-being promotion. Moreover, results based on a Chinese adolescent sample may inform youth well-being research in other developing countries globally.
Educational Investment and Adolescent Depressive Symptoms: Human Capital and Stress Process Perspectives
The influence of educational investment on adolescent mental health involves two competing theoretical pathways. Human capital theory provides a mainstream framework for understanding the positive effects of educational investment, predicting that greater resource allocation yields better cognitive and non-cognitive developmental outcomes.6 Empirical studies have demonstrated significant positive associations between family educational investment and children’s academic achievement.5,22,23
However, the stress process model challenges this optimistic prediction.24 This model posits that whether resource investment produces positive effects depends on whether the resource alleviates rather than creates new stressors. When educational investment translates into children’s feelings of financial indebtedness, time deprivation, or expectation pressure,25,26 the “stress-creation” effect may offset or even exceed the “resource-support” effect. Using data from the China Education Panel Survey, Li and Liu7 found that even after controlling for individual, family, and school characteristics, the duration of out-of-school tutoring showed a significant “depression-inducing effect” on negative emotions among adolescents. Shi et a8 further found that parental educational anxiety exacerbated adolescent depressive symptoms by increasing adolescents’ perceived stress levels.27 These findings suggest that in the domain of mental health, the protective effect of educational investment is limited and may even be counterproductive. Based on this reasoning, the present study proposes:
H1a (Material Pathway Effect Hypothesis): Family educational investment has no significant direct effect on adolescent depressive symptoms.
Hope Theory’s Protective Mechanism: Future Orientation and Adolescent Depression
In contrast to the complex effects of educational investment, the influence of future orientation on adolescent depression has a theoretically clear directional expectation. Hope theory defines hope as a positive motivational state in which pathways thinking enables individuals to generate alternative solutions when encountering obstacles, while agency thinking provides psychological energy for sustained action.13 Together, these processes allow individuals with high future orientation to reinterpret academic stress as a necessary step toward long-term goals rather than an externally imposed burden. This cognitive reappraisal constitutes an effective stress coping strategy.28
Research has found that adolescents’ hopeful future expectations are associated with higher levels of mental health.15 The longitudinal study by Ciarrochi et al18 further demonstrated a bidirectional relationship between hope and negative emotions, with hope reducing negative emotions. From the stress process model perspective, future orientation—as a psychological resource—aligns closely with the coping resources defined by Thoits,24 helping adolescents buffer environmental stress and reduce depression risk, thus serving as a crucial internal protective factor for adolescent mental health and long-term well-being. Similarly, sense of purpose in life buffers the negative impact of stress on well-being.29 In the Chinese cultural context, although Guo et al30 focused on elementary school students, their research revealed the core role of psychological variables such as expectations and confidence by demonstrating that parent-child educational expectation discrepancies affect emotional well-being through academic self-efficacy. Therefore, the present study proposes:
H1b (Psychological Pathway Protection Hypothesis): Future orientation has a significant negative effect on adolescent depressive symptoms, such that adolescents with higher future orientation have lower levels of depressive symptoms.
The Mediating Role of Depressive Symptoms: Psychological Transmission from Future Orientation to Well-Being
Future orientation may be associated with adolescent mental health not only through its association with lower depressive symptoms but also indirectly with higher well-being via reduced depressive symptoms. Keyes’ mental health continuum model31 indicates that mental health is not merely the absence of illness but a state of complete well-being. In this model, depressive symptoms and well-being represent negative and positive psychological dimensions that are correlated but distinct. High levels of depressive symptoms impair individuals’ capacity to experience positive emotions, a finding supported by extensive empirical research.28 Integrating the above theories, future orientation influences well-being precisely through the mediating variable of depressive symptoms. That is, future orientation, as a positive cognitive-motivational resource, first helps reduce adolescents’ depressive symptoms to buffer stress, and then, due to reduced depressive symptoms, releases their potential to experience well-being. Based on this, the present study proposes:
H2: Depressive symptoms have a significant negative effect on adolescent well-being.
H3: Depressive symptoms mediate the relationship between future orientation and adolescent well-being.
Comparison of Psychological and Material Pathways
Extending the logic of human capital theory and hope theory to the comparison of specific indirect effects, psychological resources (future orientation) may theoretically have stronger protective effects than material resources (educational investment). This expectation is supported on three grounds. First, regarding pathway indirectness: the effect of educational investment on mental health requires transmission through complex processes such as parent-child interaction quality and children’s subjective perceptions, whereas future orientation, as an internal cognitive appraisal, directly influences individuals’ interpretation and coping with stressors.32 Second, regarding diminishing marginal returns: in the context of intensifying educational “involution”, family educational investment may have entered a stage of extremely low marginal returns, with some studies finding “depression-inducing effects” of educational investment.7 The “suffocation effect” of normative pressure proposed by Mueller and Abrutyn11 provides a macro-sociological perspective for understanding the negative consequences of excessive investment. Third, regarding sustainability and universality: educational investment depends heavily on family economic conditions, whereas the cultivation of future orientation is less dependent on family economic circumstances and has greater cross-situational stability and malleability.33 Clarifying the relative contributions of these two types of resources to adolescent emotions and well-being can provide new mechanistic insights for research on child and adolescent well-being.
Based on the above analysis, the present study proposes the following hypotheses:
H4: Family educational investment is positively associated with adolescent well-being by alleviating depressive symptoms.
H5: Future orientation is positively associated with adolescent well-being by alleviating depressive symptoms.
On this basis, the present study constructs a parallel multiple mediation model (Figure 1) that places psychological and material resources within the same framework to analyze which pathway has stronger explanatory power for adolescent well-being.
Figure 1.

Parallel multiple mediation model diagram.
The Present Study
Figure 1 presents the parallel multiple mediation model. The model simultaneously tests two parallel pathways with depressive symptoms as the mediator: the psychological pathway from future orientation to well-being via depressive symptoms and the material pathway from educational investment to well-being via depressive symptoms. The mode determines which pathway predominates by comparing the difference between the two indirect effects.
Materials and Methods
Data Source and Sample
The present study used data from the 2022 China Family Panel Studies (CFPS), conducted by the Institute of Social Science Survey at Peking University. The CFPS is a national, comprehensive longitudinal social survey covering 25 provinces in China, representing approximately 95% of the Chinese population.34
The sample was restricted to adolescents aged 10–19 years, consistent with the World Health Organization’s definition of adolescence. By integrating individual self-report questionnaires and family economic questionnaires from the CFPS database, individuals meeting the age criteria were selected. After excluding participants with missing data on core demographic and outcome variables, the final analytic sample comprised 1151 adolescents. Missing data on educational investment were handled using multiple imputation (detailed in Data Analysis). Sampling weights were not applied in the present analysis, and the corresponding representativeness limitation is discussed in the Limitations section.
The mean age of the sample was 16.02 years (SD = 2.25), ranging from 10 to 19 years. The sample comprised 628 males (54.6%) and 523 females (45.4%). Urban household registration (hukou) accounted for 481 participants (41.8%), and rural household registration for 670 participants (58.2%).
Variable Measurement
Well-being, the dependent variable of this study, was measured using a two-dimensional framework distinguishing hedonic well-being from eudaimonic well-being. Hedonic well-being was assessed using three items adapted from Diener et al’ Satisfaction with Life Scale:35 “How happy do you feel?” (0–10), “How satisfied are you with your life?” (1–5), and “How good are your relationships with others?” (1–5). Items with different scale ranges were standardized before summation. In the present sample, the Cronbach’s α =0.79. Eudaimonic well-being was assessed using one item adapted from Ryff’s Psychological Well-Being Scale:36 “Does your life feel meaningful?”(0–10), with higher scores indicating more meaningful life.
Educational investment was one of the core independent variables. Following the operationalization method of Li and Liu,7 family annual total expenditure on children’s education was used as the indicator of family educational investment, covering monetary inputs such as tutoring fees, learning materials, and after-school care. Because the variable was right-skewed, natural logarithm transformation was applied in regression analyses [ln(educational investment + 0.01)].
Depressive symptoms was the core mediator variable. The 8-item Center for Epidemiologic Studies Depression Scale (CES-D) short form from the CFPS 2022 questionnaire was used:30 “I felt depressed”; “I felt that everything I did was an effort”; “My sleep was restless”; “I was happy”(reverse-coded); “I felt lonely”; “I enjoyed life” (reverse-coded); “I felt sad”; “I felt that life could not continue.” Each item was scored on a 4-point scale, summed to a total score ranging from 8 to 32, with higher scores indicating more severe depressive symptoms.37 In the present sample, Cronbach’s α was 0.84. Additionally, confirmatory analysis using the official CESD8 score from the CFPS data showed a correlation of 0.98 between the self-constructed depression total score and the official CESD8 score. The present study ultimately used the official CESD8 score as the operational indicator of depressive symptoms.
Future orientation was the other core independent variable. It was measured using the CFPS 2022 item “How confident are you about your future?” (1–5), with higher scores indicating greater confidence about the future. This measurement captures the cognitive-evaluative component of future orientation and serves as a proxy for the pathways thinking dimension in Snyder’s hope theory.13 However, it does not fully capture the agency thinking dimension, and this limitation is acknowledged in the Discussion.
Control variables included sex (male = 1, female = 0), age (continuous variable), household registration (urban = 1, rural = 0), and family income (annual total family income, natural logarithm transformed).
Data Analysis
All statistical analyses were conducted using R software version 4.6.0.
Missing data handling. The educational investment variable had missing values (1.74% missing before imputation), and missingness may have been associated with non-random factors such as family economic status. Multiple imputation was used to handle missing values. The R package mice was used to generate five complete datasets. The imputation model included depression, hedonic well-being, future orientation, educational investment (log), age, sex, urban/rural residence, and family income (log). Results from all datasets were pooled to obtain final estimates. Listwise deletion results are also reported as robustness checks.
Common method bias. Harman’s single-factor test was conducted, and the first principal component accounted for 30.09% of the variance (below the 40% threshold), which did not indicate a dominant single factor, although shared self-report variance cannot be fully ruled out. In addition, the core independent variables (educational investment, family income) and dependent variables (well-being, depressive symptoms, future orientation) came from different questionnaires, which further reduces the risk of common method bias.
Analysis steps. First, descriptive statistics and correlation analyses were used to examine variable distributions and relationships. Second, manifest path analysis within a structural equation modeling framework was used to construct a parallel multiple mediation model, simultaneously testing two pathways: the psychological pathway from future orientation to well-being via depressive symptoms and the material pathway from educational investment to well-being via depressive symptoms. The model included age, sex, household registration, and family income as control variables. Because all variables were observed (manifest) and the tested model was fully saturated (df = 0), global fit indices (eg, CFI, TLI, RMSEA, SRMR) were not informative and are therefore not reported. Third, bias-corrected bootstrapping with 5000 resamples was used to calculate 95% confidence intervals for indirect effects;38 a confidence interval not containing zero indicated a significant indirect effect. Fourth, the difference between the two indirect effects (ind_future – ind_edu) was calculated using model constraints, and the confidence interval was used to determine whether the two pathways differed significantly. All statistical tests were two-tailed with a significance level set at α = 0.05.
Results
Descriptive Statistics and Correlations
Descriptive statistics for all core variables are presented in Table 1. The mean depressive symptom score was 12.37 (SD = 3.44), with a range of 8 to 31. According to the CES‑D8 cutoff criteria, scores of 8–15 fall within the normal range, 16–19 indicate mild depression, and 20–32 suggest moderate‑to‑severe depression. Although the sample mean was within the normal range, individual scores reached as high as 31, indicating that some participants were at high risk for depression.
Table 1.
Descriptive Statistics of Core Variables (N = 1151)
| Variable | Mean | SD | Minimum | Maximum |
|---|---|---|---|---|
| Depressive symptoms | 12.37 | 3.44 | 8 | 31 |
| Hedonic well-being | 0 | 2.29 | –9.95 | 4.21 |
| Eudaimonic well-being | 7.47 | 1.87 | 0 | 10 |
| Future orientation | 3.91 | 0.82 | 1 | 5 |
| Educational investment (log) | 8.49 | 2.08 | –4.61 | 12.2 |
| Age | 16.02 | 2.25 | 10 | 19 |
| Family income (log) | 11.28 | 0.83 | 7.6 | 14.73 |
Notes: Hedonic well-being is the sum of three standardized items. Educational investment and family income have been natural logarithm transformed.
The mean hedonic well-being score was 0.00 (SD = 2.29), ranging from –9.95 to 4.21. The mean eudaimonic well-being score was 7.47 (SD = 1.87), ranging from 0 to 10, generally in the upper-mid range of the scale. The mean future orientation score was 3.91 (SD = 0.82), ranging from 1 to 5, with some individual variation.
All correlation coefficients had absolute values below 0.6, indicating weak to moderate correlations and no serious multicollinearity (for the full Pearson correlation matrix, see Table 2). Future orientation was significantly negatively correlated with depressive symptoms (r = –0.256, p < 0.001) and significantly positively correlated with both hedonic well-being (r = 0.492, p < 0.001) and eudaimonic well-being (r = 0.358, p < 0.001). Educational investment showed near-zero correlations with depressive symptoms (r = –0.033, p > 0.05) and hedonic well-being (r = 0.013, p > 0.05), neither of which reached statistical significance. Depressive symptoms were significantly negatively correlated with hedonic well-being (r = –0.345, p < 0.001), supporting the role of depressive symptoms as a mediator. Family income was significantly positively correlated with educational investment (r = 0.078, p < 0.01), indicating that higher family income was associated with greater educational investment.
Table 2.
Correlation Matrix Among Variables
| Variable | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
|---|---|---|---|---|---|---|---|
| 1.Depressive symptoms | 1 | ||||||
| 2.Hedonic well-being | –0.345*** | 1 | |||||
| 3.Eudaimonic well-being | –0.304*** | 0.533*** | 1 | ||||
| 4.Future orientation | –0.256*** | 0.492*** | 0.358*** | 1 | |||
| 5.Educational investment (log) | –0.033 | 0.013 | 0.043 | –0.039 | 1 | ||
| 6. Age | 0.071 | –0.137 | –0.059 | –0.128 | 0.022 | 1 | |
| 7.Family income (log) | –0.037 | 0.007 | 0.031 | –0.084** | 0.078** | 0.026 | 1 |
Note: ***p < 0.001; **p < 0.01.
Parallel Multiple Mediation Model Testing
To test the hypotheses (H4 and H5) regarding the relative explanatory power of the psychological and material pathways to adolescent well-being, manifest path analysis was used to construct a parallel multiple mediation model (Figure 1). The model simultaneously estimated two pathways: the psychological pathway and the material pathway, with age, sex, household registration, and family income included as control variables. Bias-corrected bootstrapping with 5000 resamples was used to calculate confidence intervals for indirect effects.
Path Coefficients
Both unstandardized and standardized path coefficients from the parallel multiple mediation model were estimated (see Table 3 for details).
Table 3.
Path Coefficients for the Parallel Multiple Mediation Model (N = 1151)
| Path | B | β | SE | p | 95% CI |
|---|---|---|---|---|---|
| Psychological Pathway | |||||
| Future orientation → Depressive symptoms (a1) | –1.091 | –0.257 | 0.123 | <0.001 | [–1.331,–0.851] |
| Depressive symptoms → Well-being (b) | –0.153 | –0.234 | 0.018 | <0.001 | [–0.188,–0.118] |
| Future orientation → Well-being (c1) | 1.208 | 0.433 | 0.073 | <0.001 | [1.037, 1.379] |
| Material Pathway | |||||
| Educational investment → Depressive symptoms (a2) | –0.060 | –0.036 | 0.047 | 0.203 | [–0.152, 0.032] |
| Depressive symptoms → Well-being (b) | –0.153 | –0.234 | 0.018 | <0.001 | [–0.188,–0.118] |
| Educational investment → Well-being (c2) | 0.024 | 0.023 | 0.027 | 0.384 | [–0.029, 0.077] |
| Control Variables | |||||
| Age → Well-being | –0.068 | –0.067 | 0.025 | 0.007 | [–0.117,–0.019] |
| Male → Well-being | –0.114 | –0.025 | 0.115 | 0.322 | [–0.339, 0.111] |
| Urban hukou → Well-being | –0.114 | –0.025 | 0.118 | 0.334 | [–0.345, 0.117] |
| Family income → Well-being | 0.112 | 0.041 | 0.071 | 0.115 | [–0.027, 0.251] |
Note: B = unstandardized coefficient; β = standardized coefficient; SE = standard error. Bootstrapping with 5000 resamples. Well-being refers to hedonic well-being.
Psychological pathway. The path coefficient from future orientation to depressive symptoms was negative and significant (a1 = –1.091, p < 0.001, 95% CI [–1.331, –0.851]), indicating that each one-unit increase in future orientation was associated with an approximately 1.09-point decrease in depressive symptoms. The path coefficient from depressive symptoms to well-being was negative and significant (b = –0.153, p < 0.001, 95% CI [–0.188, –0.118]), indicating that each one-unit increase in depressive symptoms was associated with a 0.15-unit decrease in standardized well-being.
Material pathway. The path coefficient from educational investment to depressive symptoms was negative but non-significant (a2 = –0.060, p = 0.203, 95% CI [–0.152, 0.032]), indicating that educational investment had no significant effect on depressive symptoms. The path coefficient from depressive symptoms to well-being was as reported above.
Control variables. Age had a significant direct effect on well-being (b = –0.068, p < 0.01), indicating that older adolescents had lower well-being. The control effects of sex, household registration, and family income did not reach statistical significance.
Model explanatory power. The model explained 7.4% of the variance in depressive symptoms (R2 = 0.074) and 13.4% of the variance in well-being (R2 = 0.134).
Indirect Effects and Comparison
The indirect effects for both pathways and the difference test between them (see Table 4) represent the core findings of this study.
Table 4.
Indirect Effects and Difference Test for the Parallel Multiple Mediation Model
| Indirect Effect | Effect | SE | 95% CI | Conclusion |
|---|---|---|---|---|
| Psychological pathway: Future orientation → Depressive symptoms → Well-being | 0.167 | 0.034 | [0.107, 0.238] | Significant |
| Material pathway: Educational investment → Depressive symptoms → Well-being | 0.009 | 0.006 | [–0.002,0.023] | Non-significant |
| Difference between pathways (Psychological – Material) | 0.157 | 0.033 | [0.101,0.231] | Significant |
Note: Bootstrapping with 5000 resamples. All models controlled for age, sex, household registration, and family income.
Psychological pathway indirect effect. The indirect effect of future orientation on well-being via depressive symptoms was 0.167 (bootstrapped 95% CI [0.107, 0.238]), with a confidence interval not containing zero, indicating a significant mediating effect. The indirect effect accounted for 12.15% of the total effect of future orientation on well-being.
Material pathway indirect effect. The indirect effect of educational investment on well-being via depressive symptoms was 0.009 (bootstrapped 95% CI [–0.002, 0.023]), with a confidence interval containing zero, indicating a non-significant mediating effect.
Difference between the two pathways. To test whether the two indirect effects differed significantly, the difference between them was calculated using model constraints. Results showed a significant difference between the two pathways (Δ = 0.157, bootstrapped 95% CI [0.101, 0.231]), with the confidence interval not containing zero. This finding is consistent with theoretical expectations: the psychological pathway effect was significantly stronger than the material pathway effect.
The path coefficients and statistical significance for each pathway in the parallel multiple mediation model are visually presented in Figure 2. As shown in the path diagram, the psychological pathway was significantly supported, the material pathway was not supported, and the two indirect effects differed significantly (Δ = 0.157, 95% CI [0.101, 0.231]).
Figure 2.

Path diagram for the parallel multiple mediation model.
Note: ***p < 0.001; n.s. p > 0.05; control variables: age, sex, household registration, family income; bootstrapping with 5000 resamples. Path coefficients are standardized. The arrows represent estimated statistical associations, not causal directions.
Robustness Checks
First, missing data handling method check. The main analysis used multiple imputation to handle missing values in educational investment. The same parallel multiple mediation model was re-run using listwise deletion (excluding cases with missing educational investment). Results showed core conclusions consistent with the main analysis: the psychological pathway indirect effect was significant (0.161, 95% CI [0.098, 0.234]), the material pathway indirect effect was non-significant (0.008, 95% CI [–0.003, 0.021]), and the difference between pathways was significant (Δ = 0.153, 95% CI [0.096, 0.225]). This indicates that the study conclusions are robust to the missing data handling method.
Second, dependent variable operationalization check. The main analysis used hedonic well-being as the dependent variable. The parallel multiple mediation model was re-run using eudaimonic well-being instead of hedonic well-being. Results showed that the psychological pathway indirect effect was significant (0.116, 95% CI [0.078, 0.161]), the material pathway indirect effect was non-significant (0.006, 95% CI [–0.003, 0.015]), and the difference between pathways was significant (Δ = 0.110, 95% CI [0.065, 0.155]). The effect size in this model was slightly smaller than in the hedonic well-being model, but the core conclusion remained consistent, indicating that future orientation was associated with both types of well-being via reduced depressive symptoms.
Third, non-linear effect check. The main analysis assumed a linear relationship between educational investment and depressive symptoms. To test for potential non-linearity, a squared term for educational investment was added to the regression model for depressive symptoms. Results showed a regression coefficient of 0.023 (p = 0.342) for the squared term, which did not reach statistical significance, confirming that the linear relationship specification was appropriate.
Discussion
The present study found that after controlling for family income, age, sex, and household registration, educational investment was not significantly associated with depressive symptoms nor significantly associated with well-being. This finding suggests that the relationship between material educational investment and adolescent mental health may not be simply positive. Previous research has indicated that high-intensity educational inputs such as out-of-school tutoring may exacerbate adolescents’ negative psychological experiences and induce emotional distress. However, the present study did not observe a significant main effect of family educational investment. This discrepancy may stem from differences in sample composition or types of investment. Li and Liu7 focused specifically on junior high school students and distinguished between types of tutoring, whereas the present study used a broader indicator of annual total educational expenditure and covered the full age range of 10–19 years. Theoretically, the effect of educational investment on mental health follows a relatively indirect pathway, and its effects may be moderated or mediated by other variables such as parent-child relationship quality, school environment, and peer effects.22,30 Therefore, the non-significant main effect of educational investment does not preclude the possibility of positive effects on mental health under specific conditions, such as appropriate investment accompanied by emotional support.
The present study found that future orientation was significantly associated with both depressive symptoms and well-being. The quantitative results indicated that each one-unit increase in future orientation was associated with an approximately 1.09-point decrease in depressive symptoms and a 1.21-unit increase in well-being. These associations are consistent with the theoretical perspective that future orientation may serve as a psychological resource for adolescent mental health. The parallel multiple mediation model further revealed that future orientation was not only directly associated with well-being but also indirectly associated with well-being via reduced depressive symptoms, with the indirect effect accounting for approximately 12% of the total effect. This finding extends the application of the stress process model24 from adult work stress to adolescent developmental stress. The model indicates that stress coping resources affect mental health through two pathways: first, directly mitigating the negative impact of stressors; second, reducing the likelihood that stressful events will trigger negative psychological consequences. In the present study, future orientation operated through the second pathway, helping adolescents maintain psychological resilience when facing challenges such as educational pressure and family expectations, thereby reducing the risk of depressive symptoms and subsequently releasing well-being that might otherwise be suppressed by depression. Related research also shows a bidirectional relationship between hope and negative emotions: hope reduces negative emotions, but negative emotions also erode hope levels,18 suggesting that cultivating future orientation may yield sustained psychological benefits.
The present study observed an asymmetry between the two mediation pathways. The parallel multiple mediation analysis results showed that the psychological pathway from future orientation to well-being via depressive symptoms was significantly supported, whereas the material pathway from educational investment to well-being via depressive symptoms did not reach statistical significance, and the difference between the two indirect effects was statistically significant (Δ = 0.157, 95% CI [0.101, 0.231]). This asymmetry suggests that psychological resources may be more strongly associated with adolescent well-being than a single material indicator (educational expenditure) in this sample. However, this finding should not be interpreted as evidence that psychological resources are universally superior to material resources. Educational investment is a multidimensional construct, and annual expenditure captures only one aspect.
The present study makes three contributions. First, it extends the discussion of human capital theory6 by suggesting that the mental health returns to educational investment may not be direct and may depend on psychosocial factors. The null association between educational expenditure and depressive symptoms does not invalidate human capital theory; rather, it highlights the need to consider non-monetary and contextual factors. Second, it provides preliminary evidence for future orientation as a correlate of adolescent mental health, although the single-item measure limits strong claims about hope theory.13 Third, it integrates the stress process model24 with positive psychology by exploring a potential association from psychological resources to well-being via depressive symptoms.
From a practical perspective, the findings tentatively suggest that targeting adolescents’ future orientation may be a promising direction for mental health promotion. However, because no intervention was tested, this implication remains speculative. School mental health programs could consider helping adolescents develop goals and meaning in learning and life, but such efforts should be evaluated in future intervention studies. Additionally, the present study found no significant association between future orientation and educational investment. This finding indicates that the two represent relatively independent resource types, supporting their placement within the same parallel mediation framework without concerns about multicollinearity or conceptual overlap.
Limitations
The present study has several limitations. First, the cross-sectional data cannot establish strict causal timing, making it difficult to rule out reverse causality: adolescents with depressive symptoms may be more likely to develop negative future perceptions, and low well-being may also exacerbate negative emotions.19 Therefore, the mediation pathway must be interpreted as a statistical association rather than a causal mechanism. Future research could use multi-wave longitudinal designs to examine dynamic relationships among variables. Second, future orientation was measured with a single item that could not distinguish between pathways thinking and agency thinking as distinct dimensions. This limits the ability to test hope theory comprehensively. Future research could use multidimensional scales to analyze the differential mechanisms of specific dimensions. Third, educational investment was measured using a total expenditure indicator that does not capture parental time, expectations, school quality, or other dimensions of educational investment, making it difficult to identify heterogeneous effects of different types of investment. Fourth, the CES-D8 scale contains items assessing positive affect (eg, “I was happy” and “I enjoyed life”), which may conceptually overlap with the hedonic well-being measure. This overlap may inflate the b path from depressive symptoms to well-being and, consequently, the estimated indirect effect, particularly for hedonic well-being. Future studies should use completely independent measures of depression and well-being to avoid criterion contamination. Fifth, eudaimonic well-being was assessed with a single item, which may not fully capture this construct. Future research could use multidimensional well-being measures. Sixth, the sample came exclusively from China, limiting cross-cultural generalizability of the conclusions. Future research could conduct cross-national comparative studies to test external validity. Moreover, although the CFPS is a nationally representative survey, the present study did not apply sampling weights, and the relatively small proportion of urban participants (41.8%) may further limit the generalizability of the findings to the broader Chinese adolescent population.
Conclusions
At a time when family educational investment is continuously increasing, the present study found that a single material indicator—annual educational expenditure—was not significantly associated with adolescent depressive symptoms or well-being. In contrast, future orientation, as a psychological resource, was significantly associated with lower depressive symptoms and higher hedonic well-being. However, owing to the cross-sectional design and measurement limitations (eg, a single-item measure of eudaimonic well-being), the results should be interpreted as statistical associations rather than causal pathways. They do not justify broad claims that challenge human capital theory or provide strong support for hope theory; rather, they suggest that psychological and material pathways may differ in their associations with adolescent mental health. Future longitudinal and, where feasible, experimental or intervention research using independent measures is needed.
Acknowledgments
The authors thank the Institute of Social Science Survey at Peking University for conducting the China Family Panel Studies and making the data publicly available. The authors also thank the participants of the CFPS for their contributions to this research.
Funding Statement
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Abbreviations
CFPS, China Family Panel Studies; CES-D, Center for Epidemiologic Studies Depression Scale; CI, Confidence Interval; SD, Standard Deviation; UNICEF, United Nations Children’s Fund.
Data Sharing Statement
The data that support the findings of this study are publicly available from the China Family Panel Studies (CFPS) at https://opendata.pku.edu.cn/dataverse/CFPS. Access to the data requires registration and adherence to the CFPS data use agreement. The authors confirm that they did not have any special access privileges to the data and that other researchers can access the data in the same manner.
Ethics Approval
The China Family Panel Studies (CFPS) was approved by the Institutional Review Board of Peking University (IRB00001052-14010). This study used de-identified, publicly available secondary data from the CFPS. According to Article 32 (Items 1 and 2) of the Measures for Ethical Review of Life Science and Medical Research Involving Human Subjects (issued on February 18, 2023, China), research using publicly available de-identified data is exempt from additional ethics review. Therefore, no further ethics approval was required for this study.
Informed Consent
All procedures involving human participants were conducted in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments. Written informed consent was obtained from all adult participants and from the parents or legal guardians of minor participants, with assent obtained from minor participants themselves. The present study used de-identified, publicly available secondary data from the CFPS and was exempt from additional ethics review.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.UNICEF. The state of the world’s children 2021: on my mind – promoting, protecting and caring for children’s mental health. United Nations Children’s Fund; 2021. [Google Scholar]
- 2.Fu XL, Zhang K, Chen XF, Chen ZY. China National Mental Health Development Report (2019–2020). Social Sciences Academic Press; 2021. Chinese. [Google Scholar]
- 3.Xiang S, Zhao S, Xiao J, et al. Longitudinal trajectories of social mobility beliefs among Chinese adolescents: the protective roles of parental academic involvement and adolescent future orientation. J Youth Adolesc. 2024;53(8):1903–13. doi: 10.1007/s10964-024-01984-x [DOI] [PubMed] [Google Scholar]
- 4.Liu A, Xie Y. Influences of monetary and non-monetary family resources on children’s development in verbal ability in China. Res Soc Stratif Mobil. 2015;40:59–70. doi: 10.1016/j.rssm.2015.02.003 [DOI] [Google Scholar]
- 5.Liu Z, Hu J, Yang Q, Qiang Q, Hou X, Wang Z. Reciprocal relationship between parental educational involvement and depressive symptoms among adolescents: a four-year longitudinal study. J Adolesc Health. 2025;77(5):900–907. doi: 10.1016/j.jadohealth.2025.07.011 [DOI] [PubMed] [Google Scholar]
- 6.Becker GS. Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education. University of Chicago Press; 1964. [Google Scholar]
- 7.Li SY, Liu AY. “Depressed children”: can out-of-school tutoring cause negative emotions? A causal inference based on two-period data from the China Education Panel Survey (EPS). Society. 2022;42(2):60–93. Chinese. doi: 10.15992/j.cnki.31-1123/c.2022.02.002 [DOI] [Google Scholar]
- 8.Shi LJ, Lin ZT, Guo HM, et al. Effect of parental educational anxiety on depression in junior high school students: mediating role of perceived stress and moderating role of parental educational time investment. J Cent South Univ Med Sci. 2025;50(9):1503–1514. Chinese. doi: 10.11817/j.issn.1672-7347.2025.250178 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Potočár L, Kozák M, Bøe T, von Soest T. The social gradients in mental health and psychosocial well-being from adolescence to midlife and the mediating role of parenting practices: a national cohort study. Soc Sci Med. 2025;383:118405. doi: 10.1016/j.socscimed.2025.118405 [DOI] [PubMed] [Google Scholar]
- 10.Gao M, Liu WF. Exploring family functioning and adolescent academic anxiety: emotional stability and social support as mediators. Psychol Res Behav Manag. 2025;18:1111–1124. doi: 10.2147/PRBM.S508537 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mueller AS, Abrutyn S. Adolescents under pressure: a new Durkheimian framework for understanding adolescent suicide in a cohesive community. Am Sociol Rev. 2016;81(5):877–899. doi: 10.1177/0003122416663464 [DOI] [Google Scholar]
- 12.Nurmi JE. How do adolescents see their future? A review of the development of future orientation and planning. Dev Rev. 1991;11(1):1–59. doi: 10.1016/0273-2297(91)90002-6 [DOI] [Google Scholar]
- 13.Snyder CR. Hope theory: rainbows in the mind. Psychol Inq. 2002;13(4):249–275. doi: 10.1207/S15327965PLI1304_01 [DOI] [Google Scholar]
- 14.Tian Y. Adolescent addictive behaviors and subjective well-being: the mediating role of mental health problems. Psychol Res Behav Manag. 2025;18:1871–1882. doi: 10.2147/PRBM.S536362 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Schmid KL, Phelps E, Lerner RM. Constructing positive futures: modeling the relationship between adolescents’ hopeful future expectations and intentional self regulation in predicting positive youth development. J Adolesc. 2011;34(6):1127–1135. doi: 10.1016/j.adolescence.2011.07.009 [DOI] [PubMed] [Google Scholar]
- 16.Ling Y, Huebner ES, Fu P, Zeng Y, He Y. A person-oriented analysis of hope in Chinese adolescents. Pers Individ Dif. 2016;101:446–450. doi: 10.1016/j.paid.2016.06.048 [DOI] [Google Scholar]
- 17.Hamilton JL, Connolly SL, Liu RT, Stange JP, Abramson LY, Alloy LB. It gets better: future orientation buffers the development of hopelessness and depressive symptoms following emotional victimization during early adolescence. J Abnorm Child Psychol. 2015;43(3):465–474. doi: 10.1007/s10802-014-9913-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ciarrochi J, Parker P, Kashdan TB, Heaven PCL, Barkus E. Hope and emotional well-being: a six-year study to distinguish antecedents, correlates, and consequences. J Posit Psychol. 2015;10(6):520–532. doi: 10.1080/17439760.2015.1015154 [DOI] [Google Scholar]
- 19.Tang P, Kostyrkaallchorne K, Butura A, Phillipsowen J, Sonugabarke E. Reciprocal developmental pathways between future-related thinking and symptoms of adolescent depression and anxiety: a systematic review and meta-analysis of longitudinal studies. Clin Psychol Rev. 2024;112:102465. doi: 10.1016/j.cpr.2024.102465 [DOI] [PubMed] [Google Scholar]
- 20.Rasmussen P, D’warte J, Gannon S, et al. “Reworldings”: exploring perspectives on the future from Danish and Australian youth during COVID-19. J Youth Stud. 2023. doi: 10.1080/13676261.2023.2206951 [DOI] [Google Scholar]
- 21.Dyson J, Jeffrey C, Third A, Threadgold S. Viable youth futures. J Youth Stud. 2025;28(3):846–849. doi: 10.1080/13676261.2025.2472959 [DOI] [Google Scholar]
- 22.Davis-Kean PE. The influence of parent education and family income on child achievement: the indirect role of parental expectations and the home environment. J Fam Psychol. 2005;19(2):294–304. doi: 10.1037/0893-3200.19.2.294 [DOI] [PubMed] [Google Scholar]
- 23.Li JL, Hu YM. How to realize the expectation that “one’s child will become a dragon”? – an analysis based on the discrepancy between parental and children’s educational expectations. Sociol Stud. 2021;36(3):204–224+230. Chinese. [Google Scholar]
- 24.Thoits PA. Stress, coping, and social support processes: where are we? What next? J Health Soc Behav. 1995;35(Extra Issue):53–79. doi: 10.2307/2626957 [DOI] [PubMed] [Google Scholar]
- 25.Zhao D, Gao YJ, Wang HF. Impact of the “Double Reduction” policy on educational investment of low-income families in cities: based on survey data of 509 families in Xi’an. J Beijing Inst Educ. 2024;38(4):28–34. Chinese. doi: 10.16398/j.cnki.jbjieissn1008-228x.2024.04.004 [DOI] [Google Scholar]
- 26.Yin QL. The logic of shadow education investment in rural families: an analytical perspective based on family function. China Youth Stud. 2025;(10):92–100. doi: 10.19633/j.cnki.11-2579/d.2025.0111.Chinese [DOI] [Google Scholar]
- 27.Jia YY, Zhang C. Traditional values and adolescent depression: unraveling mediating mechanisms through self-esteem. PLoS One. 2025;20(9):e0331660. doi: 10.1371/journal.pone.0331660 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Denovan A, Macaskill A. Stress and subjective well-being among first year UK undergraduate students. J Happiness Stud. 2017;18(2):505–525. doi: 10.1007/s10902-016-9736-y [DOI] [Google Scholar]
- 29.Burrow AL, Agans JP, Rainone N. Purpose in life as a resource for coping with stress. In: Snyder CR, Lopez SJ, editors. Oxford Handbook of Positive Psychology. 3rd ed. Oxford University Press; 2019:345–358. [Google Scholar]
- 30.Guo XL, He SRN, Qin H, Liu X. The influence of parent-child educational expectation discrepancy on elementary school students’ emotional well-being: the mediating role of academic achievement and academic self-efficacy. Psychol Dev Educ. 2019;35(4):467–477. Chinese. doi: 10.16187/j.cnki.issn1001-4918.2019.04.10 [DOI] [Google Scholar]
- 31.Keyes CLM. Mental illness and/or mental health? Investigating axioms of the complete state model of health. J Consult Clin Psychol. 2005;73(3):539–548. doi: 10.1037/0022-006X.73.3.539 [DOI] [PubMed] [Google Scholar]
- 32.Lazarus RS, Folkman S. Stress, Appraisal, and Coping. Springer; 1984. [Google Scholar]
- 33.Bandura A. Toward a psychology of human agency. Perspect Psychol Sci. 2006;1(2):164–180. doi: 10.1111/j.1745-6916.2006.00011.x [DOI] [PubMed] [Google Scholar]
- 34.Xie Y, Hu J. An introduction to the China Family Panel Studies (CFPS). Chin Sociol Rev. 2014;47(1):3–29. [Google Scholar]
- 35.Diener E, Emmons RA, Larsen RJ, Griffin S. The Satisfaction with Life Scale. J Pers Assess. 1985;49(1):71–75. doi: 10.1207/s15327752jpa4901_13 [DOI] [PubMed] [Google Scholar]
- 36.Ryff CD. Happiness is everything, or is it? Explorations on the meaning of psychological well-being. J Pers Soc Psychol. 1989;57(6):1069–1081. doi: 10.1037/0022-3514.57.6.1069 [DOI] [Google Scholar]
- 37.Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas. 1977;1(3):385–401. doi: 10.1177/014662167700100306 [DOI] [Google Scholar]
- 38.Preacher KJ, Hayes AF. Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models. Behav Res Methods. 2008;40(3):879–891. doi: 10.3758/brm.40.3.879 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are publicly available from the China Family Panel Studies (CFPS) at https://opendata.pku.edu.cn/dataverse/CFPS. Access to the data requires registration and adherence to the CFPS data use agreement. The authors confirm that they did not have any special access privileges to the data and that other researchers can access the data in the same manner.
