Roberts and colleagues provide a disciplined corrective to the habitual certainty surrounding restorative nature. Their central finding is modest but consequential: short-term exposure to natural environments produces a small reduction in depressive mood, while the evidential base remains heterogeneous, biased and methodologically fragile. The theoretical contribution is therefore not a celebration of nature’s therapeutic efficacy but a specification of the conditions under which such claims may legitimately be made. Random-effects meta-analysis, risk-of-bias instruments, GRADE assessment and machine-learning-assisted moderator selection form a methodological apparatus that tests both effect and confidence. The conceptual operation is epistemic calibration: pooled evidence is made to disclose the distance between a plausible association and a reliable causal proposition. This restraint creates a precise bridge between environmental psychology, evidence-based health research and spatial policy, showing that urban greening cannot rest on universalised benefit claims without attending to research design, population difference, environmental type and the politics of what becomes measurable.
URBANISM MEETS ART