Hi Mark,


Thanks for the explanation!

I’m looking at Montgomery county as an example, but had trouble duplicate the results.  I’m assuming that the trip distance filter was the cause.  I wonder how you apply the trip distance filter?  Is it defined by distance between centroids of residence and workplace tracts?





From: ctpp-news-bounces@chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Mark Evans
Sent: Thursday, May 26, 2016 3:34 PM
To: ctpp-news@chrispy.net
Subject: [CTPP] ACS Commute Visualization


Hey - hope this goes through. I'm Mark Evans and I developed the commuter map. Ed Christopher told me about this listserv. I'm happy to answer any questions that you may have.


I think Penelope mentioned it elsewhere, but the data is from the 2006-2010 ACS tract-tract commuter data.


I calculated a CV using the estimate and the MOE. I've seen some warnings about the high level of error in some of the ACS data with recommendations to include only data with CV < 30-40%. Unfortunately, this removes almost all the data, so the default setting I used is 80%. Even at this high level, it excludes a portion of the estimated commuters at the tract level. In addition, anyone who works and lives in the same tract would not be included.


Thanks for taking a look!