Problems have been identified with the use of data from the Australian census to estimate cycling mode share.
Travel mode to work data from the Census is regularly used to analysis and report on bike riding trends, but the usefulness of these reports is often challenged.
In a new paper Professor Jake Olivier of the School of Mathematics and Statistics at the University of NSW, and fellow researchers have found weaknesses in the Census methods and have questioned the appropriateness of using the data to estimate cycling numbers.
"The use of this data is problematic for several reasons including: (1) single day observations in winter with sampling only every five years, (2) month of data collection changed when bicycle helmet laws were introduced, (3) not possible to identify a primary travel mode, (4) the 1976 data was not a census, and (5) representing the data as a proportion can hide temporal patterns,” there authors conclude.
"Accurate cycling data is important for health and infrastructure planning.
"Cycling exposure data helps us better understand the size of bicycle-related injury/fatality by helping explain changes in injury patterns that are not due to changes in injury risk.
"Better cycling data could be used to justify increased cycling infrastructure expenditures for areas with increased cycling.”
The authors conclude:"Australia needs to collect high-quality mobility data using a standard methodology on an annual basis.
"This data is vital to our understanding of how to make transport safer and to inform policymakers where often scarce resources should be applied”.