http://disqus.com/by/sporobolus/enMon, 23 Feb 2009 13:53:40 -0000http://designaday.tumblr.com/post/80381181#comment-6502494<p>first of all this is very interesting work; i took a class from Tony Robinson in Denver (GIS in Political Science) which really focused on this basic method, and i think it is one of the more powerful ways to bring certain kinds of information across</p><p>but when i read your statement "it seems very likely ...", i shudder a bit; this type of presentation is twice removed from showing causation; first, the data underlying the map may show correlation, but that in itself proves nothing; second, and perhaps most importantly to your area of study, while choosing a spatial presentation like this can be a tremendous help for viewers to quickly and intuitively absorb information, it is much less precise than a statistical analysis of the data mapped; in other words, use maps (and other devices) to help people grapple with information, but use statistical methods to show correlation, and use deeper investigation such as the experimental method to show cause and effect</p><p>a book you might consider assigning your students is How to Lie with Maps, by Mark Monmonier</p>sporobolusMon, 23 Feb 2009 13:53:40 -0000