Hi Frank-- Hope the new year is treating you well. You asked what you're missing: How is the 3-year or 5-year window *not* like a moving average? Some of the moving parts changing from year to year are the control totals and, as result, the weights (expansion factor) that would be attached to any given household found in the 5 annual sub-samples that comprise the period sample. The population control totals come an exogenous, completely-outside-the-survey-itself, annual estimation model. And in our region, Mpls-St Paul, CB badly underestimated the top-line population totals during early and mid-decade. I think they realized it (oops), and they've been trying to make up ground ever since. Also pertaining to control totals, there may be a time-series disruption for small areas with the vintage 2009 data. CB transitioned from having county-level control totals to having sub-county control totals. And that's likely to cause weird time-series breaks for some cities where population was previously under- (or over-)estimated. ... Expect more fun this Fall, when vintage 2010 ACS are re-benchmarked to 2010 Census Enumeration ________________________ Todd Graham Principal Forecaster Metropolitan Council 390 Robert Street North Saint Paul, MN 55101 ph: 651/602-1322 email: todd.graham(a)metc.state.mn.us in: www.linkedin.com/in/toddgraham web: www.metrocouncil.org ________________________ ________________________________________ From: ctpp-news-bounces(a)chrispy.net [ctpp-news-bounces(a)chrispy.net] On Behalf Of Frank Lenk [FLENK(a)MARC.ORG] Sent: Tuesday, February 01, 2011 10:34 PM To: ctpp-news(a)chrispy.net Subject: RE: [CTPP] 2000 and 2006-08 work trip comparison question. Call me dense, but I don’t see how this problem is any different than working with moving averages. I mean, I do understand that this data is not an average but a period estimate. Still, the issues created by dropping the first year of the period and adding the last year as the data series moves forward in time seems the same to me. What am I not understanding properly? Frank Frank Lenk Director of Research Services Mid-America Regional Council 600 Broadway, Suite 200 Kansas City, MO 64105 www.marc.org<http://www.marc.org> 816.474.4240 flenk(a)marc.org<mailto:flenk(a)marc.org> 816.701.8237 From: ctpp-news-bounces(a)chrispy.net [mailto:ctpp-news-bounces(a)chrispy.net] On Behalf Of Patty Becker Sent: Tuesday, February 01, 2011 3:17 PM To: ctpp-news(a)chrispy.net Subject: Re: [CTPP] 2000 and 2006-08 work trip comparison question. If you see the point about the problem of comparing overlapping 3 year ACS estimates, it gets even worse with consecutive 5 year numbers, in which 4 of the 5 years are the same data. Personally, I think there's a strong argument for releasing 5 year data only twice a decade, by definition non-overlapping. We may get to that point in a couple of years. Patty Becker At 03:53 PM 2/1/2011, you wrote: Frank-I totally agree and I would like to see the statisticians get involved and think about this. People are going to compare overlapping periods and we have to get smart about how to deal with this. As for why some say we should not compare overlapping years the story goes like this. If you have 2006 to 2008 data and 2007 to 2009 data you would have two overlapping years. According to the logic you would in fact only be comparing the first year 2006 to the last year 2009 as the middle two years would cancel themselves out. While this sounds logical on the surface there has to be a way to deal with overlapping 5-year periods or it doesn't make sense to have an ACS. I am not sure if anyone would out right admit it but is anyone designing performance measures or other regional metrics calling for tracking annual multi-year period estimates at tracts or places? Frank Lenk wrote: I would also be interested in a more detailed explanation of why we should not compare overlapping period estimates. From a practical standpoint, people are going to do it anyway, especially the press. What do we need to know - other than the inherent issues of what is a statistically significant vs. insignificant change and the fact that ACS was designed to examine characteristics of the population rather than its level - to be able to correct naïve interpretations of easily calculated tract-level year-to-year differences in the estimates of things like poverty rate, unemployment rate, educational attainment rates, etc.? What is a correct interpretation of these differences? Frank Frank Lenk Director of Research Services Mid-America Regional Council 600 Broadway, Suite 200 Kansas City, MO 64105 www.marc.org<http://www.marc.org/> < http://www.marc.org<http://www.marc.org/>> 816.474.4240 flenk(a)marc.org< mailto:flenk(a)marc.org> 816.701.8237 From: ctpp-news-bounces(a)chrispy.net [ mailto:ctpp-news-bounces(a)chrispy.net] On Behalf Of Kendra Watkins Sent: Monday, January 31, 2011 10:05 AM To: ctpp-news(a)chrispy.net Subject: RE: [CTPP] 2000 and 2006-08 work trip comparison question. This is interesting. I recall hearing this guidance in the past. However on the Census website on the ACS page it specifically states; "Generally, you can compare American Community Survey (ACS) 3-year and 5-year estimates with Census 2000 data. There are differences in the universe, question wording, residence rules, reference periods, and the way in which the data are tabulated which can impact comparability." The qualifier in the second sentence addresses the conflict (different methods, time periods etc) but I can't find anywhere on the ACS pages where the Census recommends we not compare ACS to decennial Census. And when I search by subject the Journey to Work topic states that I can 'Compare' it to the 2000 Census. It does specifically state not to compare overlapping years on multiyear estimates. Kendra Watkins Senior Data Analyst Mid-Region Council of Governments 809 Copper Ave. NW Albuquerque, NM 87102 Phone: (505)724-3601 Fax: (505)247-1753 Email: kwatkins(a)mrcog-nm.gov< mailto:tgaudette(a)mrcog-nm.gov> This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to which they are addressed. If you have received this email in error please notify the system manager. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. From: ctpp-news-bounces(a)chrispy.net [ mailto:ctpp-news-bounces(a)chrispy.net] On Behalf Of Weinberger, Penelope Sent: Monday, January 31, 2011 8:41 AM To: ctpp-news(a)chrispy.net Subject: RE: [CTPP] 2000 and 2006-08 work trip comparison question. The CB does not recommend comparing an ACS based data set to a CB Long Form based data set. The 2000 data represent a point in time estimate, the ACS data represent a period estimate. Furthermore, the Census Bureau recommends not comparing period estimates with overlapping years. Penelope Weinberger CTPP Program Manager AASHTO 202-624-3556 http://ctpp.transportation.org/Pages/default.aspx It's just as bad to not make a plan as to blindly follow the one you already have. From: ctpp-news-bounces(a)chrispy.net [ mailto:ctpp-news-bounces(a)chrispy.net] On Behalf Of Seidensticker, Dan Sent: Friday, January 28, 2011 4:18 PM To: (ctpp-news(a)chrispy.net) Subject: [CTPP] 2000 and 2006-08 work trip comparison question. We downloaded the2006-2008 ACS county-to-county worker flow for Dane County, Wisconsin from http://ctpp.transportation.org/Pages/3yrdas.aspx. The question we now have...can that data be compared to the county-to-county 2000 CTPP work trips to determine any statistically significant increase/decrease? If so, how would one calculate the margin of error? Dan Seidensticker GIS Specialist Madison Area Transportation Planning Board: A Metropolitan Planning Organization (MPO) City of Madison Planning Unit 121 S. Pinckney Street, Suite 400 Madison, WI 53703 Voice: 608-266-9119 Fax: 608-261-9967 Email: dseidensticker(a)cityofmadison.com< mailto:dseidensticker(a)cityofmadison.com> www.MadisonAreaMPO.org<http://www.madisonareampo.org/> < http://www.madisonareampo.org/> ------------------------------------------------------------------------ _______________________________________________ ctpp-news mailing list ctpp-news(a)chrispy.net http://www.chrispy.net/mailman/listinfo/ctpp-news -- Ed Christopher 708-283-3534 (V) 708-574-8131 (cell) FHWA RC-TST-PLN 4749 Lincoln Mall Drive, Suite 600 Matteson, IL 60443 _______________________________________________ ctpp-news mailing list ctpp-news(a)chrispy.net http://www.chrispy.net/mailman/listinfo/ctpp-news ------------------------------------------------------------------------------------------------- Patricia C. (Patty) Becker 248/354-6520 APB Associates/SEMCC FAX 248/354-6645 28300 Franklin Road Home 248/355-2428 Southfield, MI 48034 pbecker(a)umich.edu