I've made this mistake before, so watch out: the new poll from the University of Iowa (released later this a.m) is not necessarily without merit, but given the weirdness of its likely voter screen, questionnaire and sample source, it's not really comparable to other polls and therefore shouldn't be used to see or describe a trend.
Home | Atlantic FAQ | Masthead | Site Guide | Subscribe | Subscriber Help
Atlantic Store | Educational Program | Jobs/Internships | Privacy Policy | Contact Us | Advertise
Copyright © 2008 by The Atlantic Monthly Group. All rights reserved.
