Adaptation and testing of a microscale audit tool to assess liveability using Google Street View: MAPS-Liveability

Abstract Background Liveability is a complex, multifaceted concept with various definitions, but with an agreed core set of features (e.g., safety, walkability). Typically, liveability is measured at the macro-level (city or regional-level), and has been used in advocacy by local populations. However, micro-level (street-level) liveability measurements could also/alternatively be used to identify modifiable environmental features impacting health and well-being. To date, no micro-level liveability tools exist. This study investigates the reliability and rater agreement of a new micro-level audit tool designed for use with Google Street View (GSV). Methods MAPS-Liveability (GSV), was adapted from the Microscale Audit of Pedestrian Streetscapes (MAPS). This study had two phases: 1) MAPS-Liveability development (rapid literature review identifying core liveability concepts, focus groups confirming liveability concepts and tool adaptation); 2) reliability investigation (researcher agreement). Assessment was made of: total liveability; nine liveability sub-characteristics (e.g., safety, health); and 12 proxy measures of behaviour including active travel (e.g., bicycle racks, presence of bicycles in racks). Inter-rater reliability and sensitivity to change were assessed by percentage agreement, inter-class correlation coefficients (ICC) and Wilcoxon signed-ranked tests (p  Results Inter-rater reliability was excellent (ICC 0.905–0.968) for total liveability, parked cars and total number of cars (moving/parked); good (ICC 0.754–0.885) for health, sustainability, places, number of bicycle racks, bicycle rack capacity, number of bicycles in the racks (time-point 2), cyclists (time-point 2), moving cars (time-point 2) and pedestrians; and moderate (ICC 0.550–0.742) for safety, inclusivity, education, traffic/transport, pavements, roads, cyclists (time-point 1), number of bicycles in the racks (time-point 1) and moving cars (time-point 1). Conclusion MAPS-Liveability provides a reliable assessment of micro-level liveability features. MAPS-Liveability has excellent inter-rater reliability for total liveability and moderate-excellent inter-rater reliability for liveability attributes and behavioural indicators. GSV at street-level supports safe, large-scale objective data collection, and collection of historical data where primary data is unavailable.