There's a pretty good literature history on this very subject. One important thing noted by JJ Flannery as early as 1956 was that if you scale your circle areas directly proportional to your variable, map viewers underestimate the values. You have to scale the the circle areas by a particular factor.
For example, see: Flannery, J.J. (1971) "The relative effectiveness of some common graduated point symbols in the presentation of quantitative data." Canadian Cartographer 8, no. 2:96-109. What's funny is that very few GISes incorporate these psychographic techniques. -Eric -=--=---=----=----=---=--=-=--=---=----=---=--=-=- Eric B. Wolf 720-209-6818 USGS Geographer Center of Excellence in GIScience PhD Student CU-Boulder - Geography On Fri, May 8, 2009 at 10:30 AM, Michal Migurski <[email protected]> wrote: >>> On May 7, 2009, at 3:48 PM, Catherine Burton wrote: >>> >>>> Nothing says well-conveyed information like a good map: >>>> >>>> http://tipstrategies.com/archive/geography-of-jobs/ >>> >>> >>> >>> Ruh-roh: >>> http://mike.teczno.com/img/jobs-gained-jobs-lost.png >> >> you're pointing out it seems to be radius rather than area based? > > > Yeah, it really blows the 100k jobs circles way out of proportion. The time > browsing is well done, but I think that using the radius as a measure was > probably a mistake. The view at the very end of the map, March 2009, is a > serious exaggeration. > > On the plus side, Hurricane Katrina in 2005 and Detroit in 2006 particularly > stand out. It would be nice if it were possible to scrub time back & forth. > > -mike. > > ---------------------------------------------------------------- > michal migurski- [email protected] > 415.558.1610 > > > > > _______________________________________________ > Geowanking mailing list > [email protected] > http://geowanking.org/mailman/listinfo/geowanking_geowanking.org > _______________________________________________ Geowanking mailing list [email protected] http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
