06.08.2013 Views

E-C158 Cover1 - Center for Transportation Analysis - Oak Ridge ...

E-C158 Cover1 - Center for Transportation Analysis - Oak Ridge ...

E-C158 Cover1 - Center for Transportation Analysis - Oak Ridge ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

T<br />

What Is New with the 2007 Commodity Flow Survey?<br />

RONALD DUYCH<br />

Bureau of <strong>Transportation</strong> Statistics, RITA<br />

he 2005 TRB CFS (1) conference provided many excellent ideas <strong>for</strong> enhancing the 2007<br />

CFS. Several creative and practical ways to leverage the data and in<strong>for</strong>mation were<br />

presented and a variety of observations were offered <strong>for</strong> immediate short-term actions while<br />

others were more strategic. Suggestions were divided into three general categories; coverage,<br />

survey methods, and data products. The following sections summarize the suggestions that were<br />

implemented and the results of those implementations.<br />

COVERAGE<br />

First and <strong>for</strong>emost, the sample size was increased to 102,000 out of approximately 754,000<br />

establishments that were considered shippers. Although an attempt was made to capture thirdparty<br />

logistics providers (3PL), the survey captured only about 12%. Similarly, an ef<strong>for</strong>t was<br />

made through precanvassing to identify auxiliaries resulting in an estimated universe of 14,800<br />

establishments. Reporting of hazardous materials was substantially improved through the<br />

precanvass and a focus on six groupings: ammonium nitrate, ethanol, explosives, hydrogen,<br />

toxic–inhalation, and all other. Even with changes to the survey that attempted to clarify<br />

containerized–intermodal movements, the CFS did not capture this category well.<br />

SURVEY METHODS<br />

Suggestions regarding use of advanced technologies and providing a continuous survey were not<br />

practical <strong>for</strong> the 2007 CFS nor was the use of the Internet <strong>for</strong> respondent reporting. However,<br />

work has begun to provide the latter <strong>for</strong> the 2012 survey. Several activities were undertaken to<br />

improve response rate including a precanvass operation and a detailed study using a presurvey of<br />

75 establishments. The precanvass covered 80,000 establishments and was used to eliminate<br />

nonshippers and to verify contact in<strong>for</strong>mation. The presurvey study was per<strong>for</strong>med to understand<br />

steps taken by responders, identify problems that responders had in completing the survey,<br />

understand how responders interpreted instructions, and evaluate their ability to provide<br />

additional in<strong>for</strong>mation such as costs. Based on a detailed postsurvey evaluation, the top reasons<br />

<strong>for</strong> nonresponse included inability to recall in<strong>for</strong>mation, in<strong>for</strong>mation sensitivity, inability to<br />

readily access needed in<strong>for</strong>mation, and awkward question <strong>for</strong>mat. Figure 1 provides a graph of<br />

the most common nonresponse items from this study.<br />

7

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!