Uncovering the mechanism linking education expectation and suicide ideation among students in tertiary education: The mediating role of academic burnout, psychache, and hopelessness

PLOS One | https://doi.org/10.1371/journal.pone.0337831 December 18, 2025 6 / 16 Proposed analysis The study employed a structural equation modelling with the maximum likelihood estimation using the Statistical Package for Social Science 27.0 AMOS 28. We specified the model with one exogenous variable (educational expectations), four endogenous variables (academic burnout, psychache, hopelessness, suicide ideation), and nine pathways based on the postulations by the 3ST and JDC theory, resulting in an over-identified model (Fig 2). The same form (model) was tested with data for each source of educational expectation (self, parents, educator/institution, culture), respectively while data for the other variables remained constant. Testing the model for the different sources of expectations allows us to better identify how each source influences poor mental health outcomes. Multigroup analysis was performed for each model with Group 1 (Malaysia), Group 2 (Singapore), and Group 3 (Australia). Estimates of global model fit for each model include the comparative fit index (CFI), normed fit index (NFI), and goodness-of-fit statistic (GFI) values near or greater than 0.95, and a root mean square error of approximation (RMSEA) value near or less than 0.06 [52]. A pairwise parameter comparison using the critical ratio for differences on the nine specified pathways was examined. For a two-tailed test, a critical ratio difference z-score between the critical range of ±1.96 indicates no significant differ-

ence. Multigroup analysis allows us to determine cross-cultural variances in the effect of different types of expectations on poor mental health outcomes. Results Preliminary analysis Prior to analysis, the dataset was screened. Incomplete responses were treated as missing with no imputation method used and were dropped from the final dataset. Note that possible bias may occur due to deletion by not accounting for the reasons for missingness. A one-way analysis of variance was conducted to examine the differences between the four sources of expectation scores and variables across the Malaysian, Singaporean, and Australian samples (Table 2). There was a significant difference between the samples for each source of expectation. Post-hoc analysis using Bonferroni cor- rection revealed that for expectations from self, the mean for Malaysia was significantly higher than Singapore (t = 2.979, p = .009). For expectations from parents, the mean for Malaysia was significantly higher than Singapore (t = 2.484, p = .040) and Australia (t = 4.724, p < .001). For expectations from educators, the mean for Malaysian was significantly higher than Singapore (t = 2.904, p = .011) and Australia (t = 4.263, p < .001). For expectations from culture, the mean for Malaysia was significantly higher than Singapore (t = 2.929, p = .011) and Australia (t = 8.398, p < .001). Moreover, the mean expectation from culture for Singapore was significantly higher than Australia (t = 5.254, p < .001). There were no significant differences

between the samples on measures of academic burnout, psychological pain, and suicide ideation. Only hopelessness was significantly different across samples, F (2, 638) =3.641, p = .027. Post-hoc analysis using Bonferroni correction revealed that mean hopelessness was significantly higher for Malaysia than Australia (t = 2.569, p = .031). Fig 2. Structural serial mediation model with defined pathways. https://doi.org/10.1371/journal.pone.0337831.g002