Background characteristics
In total ,1005 questionnaires were distributed and 963 were completed, with a response rate of 95.82%. Of the adolescents in the sample (aged 14–20, Mage = 17.72 ± 1.95 years), 55.5% were females (Mage = 17.78 ± 1.94 years), with 44.5% were classified as being with depressive symptoms. Of the parents in the sample (aged 33–67, Mage = 46.05 ± 4.39 years), 58% were mothers (Mage = 45.40 ± 4.40 years) with an average of 9.07 ± 3.77 years of education. Table 1 displays the background characteristics of the adolescents with and without depressive symptoms.
Table 1
Background characteristics of the participants.
| Without depressive symptoms (N = 544) | With depressive symptoms (N = 419) | p value |
| n,% | n,% |
Sex (N = 963) | | | 0.138 |
Male (N = 429) | 231(42.46%) | 198(47.3%) | |
Female (N = 534) | 313(57.54%) | 221(52.7%) | |
School type (N = 963) | | | 0.947 |
High school (N = 447) | 252(16.91%) | 195(46.6%) | |
College (N = 516) | 292(83.09%) | 224(53.5%) | |
Location (N = 963) | | | 0.300 |
City (N = 424) | 247(45.4%) | 177(42.2%) | |
Town (N = 292) | 154(28.3%) | 138(32.9%) | |
Village (N = 247) | 143(26.3%) | 104(24.8%) | |
Mean Age (SD) (N = 963) | 17.739(1.944) | 17.695(1.962) | 0.677 |
Parents who spent more time taking care of the children in daily life (N = 963) | | | 0.589 |
Mother (N = 559) | 315(57.90%) | 244(58.2%) | |
Father (N = 404) | 229(42.10%) | 175(41.8%) | |
Parent mean Age (SD) (N = 963) | 46.09(4.450) | 46(4.318) | 0.849 |
Parent educational years (SD) (N = 963) | 9.433(3.617) | 9.539(3.622) | 0.509 |
p value: Chi-square test and Mann-Whitney U test |
Table 1. Background characteristics of the participants.
---------------------Table 1-------------------------
Correlations between depression and other variables
Table 2 shows the Spearman correlations between predictive variables and depression. We found significant correlations between depression and social benefits of VOC (r = 0.111, p < 0.01), relational benefits of VOC (r = − 0.08, p < 0.05), psychological benefits of VOC (r = − 0.108, p < 0.01), parenting style (r = − 0.499, p < 0.01), external LOC (r = 0.375, p < 0.01), and self-esteem (r = − 0.495, p < 0. 01).
Table 2
Spearman’s correlation between main variables.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 |
1 Sex | 1 | | | | | | | | | | | |
2 Age | 0.036 | 1 | | | | | | | | | | |
3 School type 4 Location 5 Parent age 6 Parent educational years | 0.047 0.020 − 0.65 − 0.047 | 0.915** 0.072* 0.063 0.023 | 1 0.056 0.049 0.020 | 1 − 0.008 − 0.089** | 1 − 0.064** | 1 | | | | | | |
7 Social benefits | − 0.239** | − 0.015 | − 0.034 | 0.132** | 0.029 | 0.046 | 1 | | | | | |
8 Relational benefits | − 0.131** | − 0.28 | − 0.035 | − 0.025 | 0.009 | 0.022 | 0.296** | 1 | | | | |
9 Psychological benefits | 0.015 | 0.051 | 0.035 | − 0.100* | 0.002 | 0.018 | 0.106** | 0.530** | 1 | | | |
10 Parenting style | 0.107** | 0.037 | 0.042 | − 0.118* | 0.050 | 0.038 | − 0.194** | 0.110** | 0.232** | 1 | | |
11 External LOC | − 0.110** | − 0.015 | − 0.051 | 0.032 | 0.000 | − 0.071* | 0.236** | 0.045 | − 0.067* | − 0.312** | 1 | |
12 Self-esteem | 0.071* | 0.037 | 0.048 | − 0.074* | 0.019 | 0.026 | − 0.123** | − 0.186* | 0.228** | 0.505** | − 0.379** | 1 |
13 Depression | − 0.022 | − 0.020 | 0.000 | 0.027 | − 0.004 | 0.023 | 0.111** | − 0.080* | − 0.108** | − 0.499** | 0.375** | − 0.495** |
Table 2. Spearman’s correlations between main variables.
--------------------Table 2---------------------
Multiple mediating model
We developed multiple mediating models using the following steps to assess the multiple mediator effects of parenting style, external LOC, and self-esteem in the relationship between VOC and depression.
Analysis of direct effects
We established a direct effects model of VOC on depression, and the results indicate that this model is a good fit (χ2/df = 5.599, CFI = 0.935, TLI = 0.920, RMSEA = 0.069 (0.061–0.077), SRMR = 0.034). Moreover, the direct effects of social and psychological benefits on depression were significant (βsocial benefits = 0.191, (95% CI, 0.132, 0.248), p < 0.001; βpsychological benefits = − 0.180, (95% CI, − 0.264, − 0.094), p < 0.001). However, relational benefits did not directly predict depression (βrelational benefits = − 0.027, (95% CI, − 0.104, 0.052), p > 0.05) and hence were not included in the subsequent multiple indirect analysis.
Analysis of multiple indirect effects
We developed a multiple indirect model based on the direct model (see Fig. 1). The results demonstrated that the model was acceptable (χ2/df = 5.501, CFI = 0.936, TLI = 0.914, RMSEA = 0.068 (0.062–0.075), SRMR = 0.037).
Moreover, the results showed that the social benefits of VOC indirectly predicted depression in two ways, including a one-step path involving parenting style (βsocial benefits = 0.070 (95% CI, 0.051, 0.093), p < 0.001) and an external LOC (βsocial benefits = 0.039 (95% CI, 0.025, 0.058), p < 0.001) and a two-step path involving parenting style through both an external LOC (βsocial benefits = 0.010 (95% CI, 0.006, 0.016), p < 0.001) and self-esteem (βsocial benefits = 0.033 (95% CI, 0.022, 0.045), p < 0.001). The psychological benefits of VOC indirectly predicted depression in two ways, including a one-step path involving parenting style (βpsychological benefits = − 0.085 (95% CI, − 0.113, − 0.064), p < 0.001) and self-esteem (βpsychological benefits = − 0.040 (95% CI, − 0.058, − 0.026), p < 0.001) and a two-step path involving parenting style through both an external LOC (βpsychological benefits = − 0.013 (95% CI, − 0.019, − 0.008), p < 0.001) and self-esteem (βpsychological benefits = − 0.040 (95% CI, − 0.055, − 0.028), p < 0.001) (see Table 3). The findings also showed that the social benefits of VOC predict depression positively through an external LOC but not through self-esteem; meanwhile, psychological benefits of VOC predict depression negatively through self-esteem but not through external LOC.
Table 3
Bias-corrected bootstrap test of direct and indirect effects.
Paths | Standardized β | 95% CI |
Low | High |
Social benefits | | | |
Social-depression | 0.019 | − 0.038 | 0.080 |
One-step mediation | | | |
Social-parenting།depression | 0.070*** | 0.051 | 0.093 |
Social-external -LOC།depression | 0.039*** | 0.025 | 0.058 |
Social-self-esteem།depression | 0.013 | − 0.001 | 0.030 |
Two-step mediation | | | |
Social-parenting།external -LOC།depression | 0.010*** | 0.006 | 0.016 |
Social-parenting།self-esteem།depression | 0.033*** | 0.022 | 0.045 |
Psychological benefits | | | |
Psychological-depression | − 0.009 | − 0.082 | 0.064 |
One-step mediation | | | |
Psychological-parenting།depression | − 0.085*** | − 0.113 | − 0.064 |
Psychological-external -LOC།depression | − 0.007 | − 0.018 | 0.003 |
Psychological-self-esteem།depression | − 0.040*** | − 0.058 | − 0.026 |
Two-step mediation | | | |
Psychological-parenting།external-LOC།depression | − 0.013*** | − 0.019 | − 0.008 |
Psychological-parenting།self-esteem།depression | − 0.040*** | − 0.055 | − 0.028 |
------------Figure 1-----------
Table 3. Bias-corrected bootstrap test of direct and indirect effects.
------------Table 3------------