For those of you not into academics or statistics, an ANOVA is short for Analysis of Variance. It's a statistical test to determine differences between more than two groups. A t-test determines differences between two groups. Growing up in Smackover I never gave any thought to this type of statistic, not that I knew it existed. To me, statistics were what you came up with after adding up the totals of a football game or a basketball game. 100 yards rushing on 20 carries, 12 completions on 24 pass attempts, 8 tackles, 3 for a loss. Those were the types of statistics that I grew up with. Now that I live in a different world, I am dealing with the "other" type of statistics. I am in search of a test that will tell me what the differences are once I find them.
For example, if I have five groups and I run the ANOVA and find that differences exist between the groups, I then need to find out what the differences are. Is it between group one and group four or is it group two and group five?
I've heard of Tukey's test, but I'm not sure I have a handle on that yet.
Anyone out there have a suggestion?
Friday, January 12, 2007
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment