ANOVA ques----can someone help?

#2
Saralle,

The sum of squares of treatment and error add up to total sum of squares for one-way ANOVA. We are not partitioning variances.
 
#3
variance

I am very new to statistics, so perhaps I did not use the correct term. If you find the total variance for a group of 20 subjects , the variance between families and between environments, and the variance within each of the ten pairs. There were two sets of ten pairs given, with the same numerical scores (the same 20 numbers arranged differently between environments and families. The total variance for each group came to the same number. In the first group, the VB family, VB environment, and VW the ten groups added to the total variance, and in the second it did not work out. I checked my calculations again, and still no. Any hints as to why this might be? I hope I explained it ok.
Thanks!
 
#4
Saralle,

In the Examples forum there is a post entitled "Online Statistics Resources". It has a list of entry level online statistics textbooks. You may want to read the relevant chapters.

If you are calculating variances, they will not add up. Sum of squares add up to SS total (SStotal=SStreat+SSerror), but not variances.