Age-specific associations between perceptions of community environmental quality, restorative environmental perception and community identity: an interpretative perspective based on the Homo Urbanicus

Context Understanding the associations between residents’ perceived community environmental quality and community identity, and their age-related differences, is important for advancing all-age-friendly communities. Methods Drawing on Hok-Lin Leung’s Homo Urbanicus theory, this study analyzed 929 questionnaires collected in 2025 from 36 residential communities in Yangpu District, Shanghai. Structural equation modelling was used to examine the association structure among perceived community environmental quality, perceived restorative environment and community identity, and to test whether the three structural paths differ across four age groups (children, young adults, middle-aged and older adults), using measurement invariance tests, cross-group equality constraints and paired parameter comparisons, with cluster-robust standard errors. Results (1) Perceived restorative environment significantly mediated the association between perceived community environmental quality and community identity: the indirect effect accounted for 51.5% of the total effect and the direct effect remained significant, the two links being comparable in magnitude. (2) Measurement invariance was broadly supported, permitting cross-group comparison of the structural paths, and the mediation held in all four groups (bias-corrected bootstrap, 5,000 resamples). Cross-group differences were localized: the path from perceived restorative environment to community identity was stronger in the middle-aged than in the older group (cluster-robust difference = 0.270, p = 0.021; significant under all three specifications). The children’s group scored numerically higher on the path from perceived community environmental quality to perceived restorative environment, but this did not consistently reach significance and is reported descriptively. Direct effects did not differ significantly across the four groups (Δχ 2 = 2.670, Δdf = 3, p = 0.445), so the non-significant direct effect in the children’s group is not interpreted as an absent effect. Discussion These findings are associations derived from cross-sectional data, not causal effects or observed psychological transformation. Within the ‘events, people, time and space’ framework of the Homo Urbanicus theory, life-stage differences in the weighting of self-existence and coexistence needs offer a coherent interpretation of the age-specific path patterns. The results support the differentiated provision of restorative space in age-friendly community planning.