RE: [CTPP] 2000 and 2006-08 work trip comparison question.

Comparing ACS data products to 2000 data products can get a little tricky but it is possible to get a handle on the statistical significance of the changes. The best information I have found on the subject can be found in a report done by CUTR of South Florida at http://www.nctr.usf.edu/pdf/77802.pdf I would take a look at the report before jumping to the "calculator" that they developed. When estimating the Standard Error and MOEs for 2000 data you will need to know what the Census Bureau calls the Survey Design Factor. The nice thing with the CUTR calculator is that they have embedded all the design factors for all the states into the spread sheet. Here is the link to it. http://www.nctr.usf.edu/spreadsheet/77802.xls Here is the main link to the CUTR South Florida work http://www.nctr.usf.edu/abstracts/abs77802.htm Another resource worth having around is the NCHRP report 588 which focuses on the ACS and talks a lot about statistical significance testing. It is online at http://onlinepubs.trb.org/onlinepubs/nchrp/nchrp_rpt_588.pdf Finally Liang Long who is working for FHWA can also help (liang.long(a)dot.gov) as she did all the calculations and comparisons between ACS data CTPP 2000 data that are on the CTPP data profiles found at http://ctpp.transportation.org/Pages/profiles.aspx Now for a word of CAUTION. When using the 3-year CTPP I would be careful of sub-county flows. We know that the Census Bureau changed the weighting of the ACS in 2009 to use sub-county population estimates. At a county level anything form 2008 and earlier should be OK but below county I would be careful and consider this any analysis. Hopefully as more people get a chance to dig into their 3-year 06-08 CTPP data products we will all learn more. -- Ed Christopher FHWA Resource Center Planning Team 4749 Lincoln Mall Drive, Suite 600 Matteson, IL 60443 708-283-3534 (V) 708-574-8131 (C)