Non-response has become one of the most serious problems affecting household and personal travel surveys in North America and elsewhere. Response rates typically range from around 10-15 percent for some mail-back surveys, to up to 45 percent for telephone retrieval surveys. The response rate for the NHTS was sufficiently low that OMB gave only provisional permission for the survey to proceed, citing the fact that the response rate was well below that expected of other government surveys. Low response rates have the potential to introduce substantial biases into the data collected, and may result in significant segments of the population either not being included or being severely under-represented. Project NCHRP 8-37 is focusing on standards for household and personal travel surveys in an effort to improve not only the comparability between surveys in different places and times, but also the overall quality of these surveys. A critical element of this, which was identified as being beyond the scope of the project, is the issue of survey non-response. It is the intent of this proposed research to address that critical gap. There are two primary issues to be addressed in the matter of non-response. The first is to identify reasons why people do not respond and to determine if there may be strategies or options that could be used to increase response significantly. The second is to identify who the nonrespondents are and to determine the likely magnitude of biases resulting from non-response. Both of these goals can be satisfied by a non-response survey and specialized analysis of the data both from respondents and nonrespondents. Such a survey needs to be conducted as a follow-on to an existing survey that is in the field within the time frame of this project. An opportunity to do this exists in St. Louis, where a survey conducted in the Fall of 2002, and a non-response follow-up survey could be added on to the survey period.
The purpose of this task is to conduct a non-response follow-up survey in St. Louis that will attempt to ascertain reasons for not responding to the original survey, incentives or other strategies that could have been used to improve response, and to collect some information on the demographics of the household and a summary of their travel. The research team will also conduct a literature review at the outset, both to identify what is already known about non-respondents to household travel surveys and to assist in the design of the non-response survey for this research.
For more information, please visit: http://www.trb.org/TRBNet/ProjectDisplay.asp?ProjectID=911