>From the "Information for Authors" section of _American Journal of Psychiatry_, a journal that (up until this moment) I thought I loved: > http://ajp.psychiatryonline.org/misc/ifora.shtml > > Figures > [...] > Content > > 1. Error bars should not be used. Explanation, anyone? They don't say anything about using confidence intervals instead. They want naked symbols. It seems...indecent. go figure, David Epstein [EMAIL PROTECTED]
