“…This was a cross-sectional survey, so we have insight into dynamics over time and we cannot engage in causal inference. Nor did we use experience sampling methods (see Solymosi, Bowers, & Fujiyama, 2015, 2018Chataway & Hart, 2016;Chataway, Hart, & Bond 2019). Ideally, we would have taken into account social and structural characteristics of the neighbourhood (see Brunton-Smith et al, 2014;Côté-Lussier et al, 2015) and use innovative statistical techniques to estimate cluster-level means and variance (Brunton-Smith, Sturgis, & Leckie 2018;Jackson et al, 2019).…”