Paul Brandon asked for stats mavens to address his inquiry: "How many students do you need to justify the assumption of a normal (or other) distribution of measures of performance?" -------------------------- If you know any mavens, you may suspect that they will not have a simple answer for him. :)
There isn't a particular number that will ensure a normal distribution if the population isn't normally distributed in the first place. Probably the closest a college class would come to the normal distribution of raw scores concerning knowledge of an academic subject would be in an Intro class required of all majors. Even then, distributions are often bimodal. Once you get to upper division major classes, it is very difficult to justify the expectation of a normal distribution, no matter how many students may be in a class. Of course, there are always exceptions (which is where the mavens come in...). Rick Dr. Rick Froman, Chair Division of Humanities and Social Sciences Professor of Psychology Box 3519 John Brown University 2000 W. University Siloam Springs, AR 72761 rfro...@jbu.edu<mailto:rfro...@jbu.edu> (479) 524-7295 http://bit.ly/DrFroman "The LORD detests both Type I and Type II errors." Proverbs 17:15<http://www.biblegateway.com/passage/?search=proverbs%2017:15&version=NIV> --- You are currently subscribed to tips as: arch...@jab.org. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=30450 or send a blank email to leave-30450-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu