Originally Posted by
mr_bill
...Add in noisy numbers (+/- 15% for Portland, +/- 30% and higher for Boston), and you can make whatever meaning you'd like from the numbers. The numbers are barely statistically meaningful, so make up a fact, whatever you'd like. That's the brilliance of some of the ACS numbers, like bike mode share.
The magnitude of the MOE still covers the differences some of us are citing. The semi-controlled comparison numbers I cited a page ago won't support making up facts. 2012: 4.5% +/- 0.8% MOE is larger than 2005: 2.4% +/- .5% --that's real data, and real growth. And, as I alluded to, I suspect that a lot of that MOE is month-of-year variation due to climate--the ACS is conducted every month, so Minneapolis, and other leading cycling areas, probably show lots of variance, but probably also approach or exceed double digit mode share in good weather in some areas . Yes, there is longitudinal variation. Yes, the numbers are still small. But there are data that need to be interpreted, not dismissed. I realize this is A&S, but numbers matter, too.