Emotional adjustment and peer relationships: The role of behavioural reputation and classroom social climate

PLOS One | https://doi.org/10.1371/journal.pone.0332297 September 18, 2025 6 / 13 along with self-worth, to all behavioural reputation measures and social preference at Time 1. Gender was included as a covariate, and covariances were included among all predictor variables. We then trimmed the non-significant paths. The model obtained a good fit, Χ2(9) = 15.73, p = .073, CFI = .998, RMSEA = .023, SRMR = .010, see Fig 1. After controlling for self-worth, anger was positively associated with having a reputation for both disruptive behaviour and starting fights, and negatively predicted a cooperative reputation and a shy reputation. Anxiety was positively associated with a shy reputa- tion, but negatively predicted a reputation for starting fights. Finally, Time 2 social preference was positively predicted by a cooperative reputation and negatively by a shy and a disruptive social reputation. Indirect pathways from Time 1 emotional adjustment to Time 2 social preference, via the relevant Time 1 behavioural reputation scores were estimated, using bootstrapping with 1000 samples. To ensure conservative estimates, the direct paths from Time 1 emotional adjustment to Time 2 social preference were included, even if they did not achieve statisti- cal significance. Anger predicted lower social preference via a less cooperative reputation (standardised estimate = −.04, p = .001) and a more disruptive reputation (standardised estimate = −.06, p = .001). Anger also modestly predicted greater

social preference via a less shy reputation (standardised estimate = .02, p = .012). Finally, anxiety predicted lower social preference via a more shy reputation (standardised estimate = −.03, p = .006). Our multigroup analysis compared the 20 most and 20 least socially supportive classrooms, as evaluated in terms of peer experiences. Beginning with an initial saturated model, we placed equality constraints on all paths from Time 1 behavioural reputation to Time 2 social preference, after controlling for Time 1 social preference. This showed that con- straints do not hold across groups, Χ2(5) = 15.94, p = .007. Inspection of the unconstrained saturated model showed that the negative link from disruptive behaviour to Time 2 social preference was robust in the high social climate classrooms but virtually zero in the low social climate classrooms. When the equality constraint was freed for this path, the chi-square returned to non-significant, Χ2(4) = 5.27, p = .261, ∆Χ2(1) = 10.67, p = .001. The model continued to fit well when paths from emotional adjustment to behavioural reputation were set to be equal across the two classroom social climate groups, Χ2(33) = 36.97, p = .291. The final model, with non-significant paths deleted, had excellent fit, Χ2(45) = 44.93, p = .475, CFI = .1.00, RMSEA = .000, SRMR = .026, see Fig 2. Regarding indirect effects, anger predicted lower social preference via a less cooperative reputation regardless of classroom climate (standardised estimate = −.05, p = .002). However, only

in classrooms with a higher concentration of positive peer interactions did anger predict lower social preference via more disruptive reputation (standardised estimate = −.12, p < .001); the indirect effect in classrooms with a relatively lower con- centration of positive peer interactions was non-significant (standardised estimate = −.02, p = .579). Table 1. Correlations among all variables. 1 2 3 4 5 6 7 8 9 10 1. Angry -- 2. Anxious .58** -- 3. Depressive .68** .76** -- 4. Self-worth −.34** −.26** −.36** -- 5. Cooperative −.19** −.05* −.11** .16** -- 6. Disruptive .22** .04 .11** −.13** −.37** -- 7. Shy −.05 .15** .10** −.05 .14** −.22** -- 8. Fight .21** −.03 .05 −.09** −.31** .75** −.25** -- 9. Leadership −.05 −.06* −.07** .13** .43** −.10** −.15** −.02 -- 10. Soc Pref T1 −.20** −.14** −.18** .16** .48** −.47** −.05 −.32** .31** -- 11. Soc Pref T2 −.18** −.13** −.17** .13** .39** −.39** −.05 −.29** .25** .60** * p < .05 ** p < .01. Note. Soc Pref = Social preference. Ns ranged between 1296 and 1439 depending on missing data. https://doi.org/10.1371/journal.pone.0332297.t001