The CFS methodology can and should be improved. Evaluation of the methodological changes implemented prior to the 1997 CFS indicated that these were successful in providing a more efficient sample design, lessening respondent burden, and improving timeliness of data products. In the past 10 years, survey methodology has benefited from many research-based improvements in sample design, instrument design, data collection, and nonresponse mitigation. For various reasons, the CFS has not taken advantage of such methodological improvements. It is time to do so. This paper has discussed current issues, challenges, and survey considerations that are relevant to the CFS. It underscores the need for a continuous research and evaluation plan that would enable BTS and the Census Bureau to stay abreast of methodological enhancements, as well as real changes in freight transportation, mat impact CFS data quality and usefulness. Such a research plan could provide an increased understanding of the response process and respondent burden toincrease response rates, alleviate privacy and confidentiality concerns, improve data quality, etc., for establishment surveys. This is an area in which not enough research is currently taking place. The issues discussed in this paper also draw attention to the value that ongoing dialogue with data users and stakeholders provides. Not only do these persons bring substantive expertise on tough decisions such as whether or not to change the way shipments are sampled, but also their input helps to ensure methods employed are providing data of most importance and utility.
展开▼