Fig 1.
A mediation analysis was conducted with the SAS-SV as the independent variable (X), the PSQI as the dependent variable (Y), and the PSS and the HPLP-II as the mediators M1 and M2. The coefficient c is the total effect between X and Y, and c′ is the direct effect of X on Y whilst controlling for M1 and M2. The SAS-SV demonstrated positive correlations with the PSQI and the PSS, and a negative correlation with the HPLP. We further hypothesize that The HPLP exhibited a negative correlation with the PSQI, whereas the Perceived Stress Scale (PSS) exhibited a positive correlation with it.
Table 1.
Characterization of the participants for the full samples (N = 2317).
Table 2.
Scores on various scales for college students with different characteristics.
Table 3.
Correlations between the main study variables (N = 2317).
Table 4.
Mediation analysis of SAS-SV and PSQI.
Fig 2.
Mediation model of HPLP and PSS between SAS-SV and PSQI.
The path coefficients in the parallel mediation model are delineated as follows. The coefficient a₁ signifies that the SAS-SV is a significant positive predictor of the PSS, whereas a₂ denotes its significant negative predictive effect on the HPLP-II. After controlling for SAS-SV and HPLP-II, the path coefficient b₁ indicates that the PSS remains a significant positive predictor of the PSQI. Conversely, after controlling for the SAS-SV and PSS, the coefficient b₂ shows that the HPLP-II is a significant negative predictor of PSQI. Furthermore, two specific indirect effects were identified: smartphone addiction impairs sleep quality by increasing perceived stress (a₁b₁), and similarly, by reducing health-promoting lifestyle behaviors (a₂b₂). The coefficient c is the total effect between X and Y, and c′ is the direct effect of X on Y whilst controlling for M1 and M2. ***P < 0.001.