Dan Froomkin of The Washington Post makes an excellent point about the effect of word choice in polling, one of the reasons polling stories should be handed with extreme caution. He puts up to polls with significant differences in results, and tracks back to the polls' questions to find out why they diverged. (That's not even getting to online polls, which serve no useful purpose beyond encouraging web hits. A subject for another day.)
But before you get to the poll, read his thorough assessment of the brouhaha over The New York Times' (and other papers) decision to publish information about the tracking of financial records.
No comments:
Post a Comment