Abstract
OBJECTIVE: Policies and interventions increasingly aim to reduce smoking in outdoor public spaces, but evidence on factors influencing smoking in specific locations remains limited. Systematic observation can unobtrusively assess behaviours in environmental contexts, reducing biases from self-report. This study aims to develop and test the reliability and validity of MOSMOKE (Method for Observing SMOKing and vaping bEhaviours): a new tool for assessing the number of people holding or inhaling a cigarette or vape in public spaces. DESIGN: MOSMOKE was adapted from a previously validated observation tool for assessing physical activity and well-being behaviours. Following piloting and refinement, inter-rater reliability for assessing smoking, vaping and age group classification was analysed using intraclass correlation coefficients (ICCs). A main study assessed criterion-related validity through 32 hours of observations over 4 days. A 2×2 study design was used, with four sites selected that varied by two environmental characteristics: presence of a smoking bin and adjacency to an office block. SETTING: Four public spaces in Manchester, UK. RESULTS: Inter-rater reliability was 'good' (ICCs>0.75; n=4) or 'excellent' (ICCs>0.90; n=2) for smoking behaviours, and mostly 'good' (n=4) or 'excellent' (n=1) for vaping behaviours. Observed differences in smoking and vaping behaviours across sites aligned with prespecified hypotheses that smoking would be more prevalent near smoking bins (p=0.02) and office entrances (p=0.006), supporting criterion-related validity. CONCLUSIONS: This study provides preliminary evidence that MOSMOKE is a reliable and valid tool for unobtrusively assessing smoking and vaping in public spaces. It can be used to evaluate policies and interventions targeting smoking or vaping in specific environmental contexts. MOSMOKE is freely available, with a detailed manual to support its use.