Increasing active travel: aims, methods and baseline measures of a quasi-experimental study

BACKGROUND: Policy advisers are seeking robust evidence on the effectiveness of measures, such as promoting walking and cycling, that potentially offer multiple benefits, including enhanced health through physical activity, alongside reductions in energy use, traffic congestion and carbon emissions. This paper outlines the ‘ACTIVE’ study, designed to test whether the Model Communities Programme in two New Zealand cities is increasing walking and cycling. The intervention consists of the introduction of cycle and walkway infrastructure, along with measures to encourage active travel. This paper focuses on the rationale for our chosen study design and methods. METHOD: The study design is multi-level and quasi-experimental, with two intervention and two control cities. Baseline measures were taken in 2011 and follow-up measures in 2012 and 2013. Our face-to-face surveys measured walking and cycling, but also awareness, attitudes and habits. We measured explanatory and confounding factors for mode choice, including socio-demographic and well-being variables. Data collected from the same households on either two or three occasions will be analysed using multi-level models that take account of clustering at the household and individual levels. A cost-benefit analysis will also be undertaken, using our estimates of carbon savings from mode shifts. The matching of the intervention and control cities was quite close in terms of socio-demographic variables, including ethnicity, and baseline levels of walking and cycling. DISCUSSION: This multidisciplinary study provides a strong design for evaluating an intervention to increase walking and cycling in a developed country with relatively low baseline levels of active travel. Its strengths include the use of data from control cities as well as intervention cities, an extended evaluation period with a reasonable response rate from a random community survey and the availability of instrumental variables for sensitivity analyses.