Design considerations for field studies of changes in indoor air pollution due to improved stoves

Monitoring the actual improvements in indoor air quality as a result of installation of improved stoves is critical in assessment of the effectiveness of improved stove interventions. Households are complicated places, however – many things that affect air pollution are the same in households within a geographically distinct locality, but there are many factors that are different. Thus, to be sure that any air pollution differences that might be measured in households with different types of stoves are really attributable to true differences and are not merely due to natural variations unrelated to the stoves, it is necessary to follow appropriate sampling designs with sufficient numbers of measurements, or a large enough sample size. This paper presents a summary of cross-sectional and before-and-after statistical designs to evaluate the effectiveness of an improved biomass or other solid-fuel stove as the best indication of how much benefit has actually been achieved by the stove in practice in households under normal conditions. It also highlights considerations when selecting one design over another, and key considerations in lowering the samples sizes required to demonstrate statistically significant differences by reducing variation. An interactive website using Excel functions is available for facilitating calculations.