Make three groups of three columns on a sheet of paper with as many rows as your experiment has samples. The columns will be labeled "name of condition," "x - xbar," and "²."
Write the values for each sample in the first row of the column one to a row. At the bottom of each column, calculate xbar for each group identified as sub 1, sub 2, and sub 3. Xbar is the mean of the x. This is assuming there are three levels. Then calculate xbar bar, the grand mean or the mean or means. In the second column, calculate x minus xbar. In the third column, square the differences. At the bottom, calculate the sum of squares (SS) sub 1, sub 2, and sub 3.
Here's an example.
X1- 6, 7, 8 xbar is 7
X2- 9, 1, 5 xbar is 5
X3- 1, 3, 5 xbar is 3
xbar bar is 5
6-7=-1 | 1 |
7-7=0 | 0 |
8-7=1 | 1 |
The sum of squares is 2 for group one.
Continue the same for groups two and three.
Create a source table by drawing a five column by four row table. The columns will be "source," "SS," "df," and "F." The sources will be between, within and total. Fill in the table with the information from the previous work table. Remember that SS between groups is equal to the sum of the difference of xbar and xbar bar, squared times n.