A further limitation was that all the measures were self-reported and web-based, which could lead to participants exaggerating their actual physical activity efforts and healthy eating behaviour. It would have been useful to compare individual pedometer steps to the self-reported data, but this information was not available through the 10,000 steps program. Web surveys can produce lower response rates and poorer data quality than a face-to-face survey.31 This may mean that respondents differentiated less on the rating scales (i.e. chose the same response for different questions)31 that were used for some of the process evaluation questions….