So you think you understand how to read error bars in graphs, right? You may well think that they are extremely important to the proper interpretation of graphs. I bet many of you teach your students how to interpret error bars, and even insist that your students use error bars in their own graphs. Now, try these two problems: http://scienceblogs.com/cognitivedaily/2008/07/ill_bet_you_still_dont_underst.php . . . . . . . . . . . Have you done it? Good. You are then permitted to read another post from more than a year ago (when the same blogger asked roughly the same question) where many of you, in all probability, will learn that you have been misinterpreting the meaning error bars all this time, and have been misteaching it to your students. (Hint, error bars that don't overlap *don't* actually demonstrate a significant difference between the points. You may stop screaming now and read the explanation: http://scienceblogs.com/cognitivedaily/2007/03/most_researchers_dont_understa.php
"But Chris!," I hear you cry, "This is just a stupid blog. Why should I believe him?" Because it is based on a 2005 article in /Psychological Methods/ (see the bottom of the blog entry for the full citation). Chris -- Christopher D. Green Department of Psychology York University Toronto, ON M3J 1P3 Canada 416-736-2100 ex. 66164 [EMAIL PROTECTED] http://www.yorku.ca/christo/ "Part of respecting another person is taking the time to criticise his or her views." - Melissa Lane, in a /Guardian/ obituary for philosopher Peter Lipton ================================= --- To make changes to your subscription contact: Bill Southerly ([EMAIL PROTECTED])
