Original Post
Suppose I have an n x n grid of real numbers. I want to do a variance analysis on every 2 x 2 area. I could do it with one of 2 methods 1. Each 2 x 2 area has unique group of cells (0,0) (0,1) | (0,2) (0,3) ... (1,0) (1,1) | (1,2) (1,3) ... (2,0) (2,1) | (2,2) (2,3) ... (3,0) (3,1) | (3,2) (3,3) ... ... .... 2. Overlapping 2 x 2 areas (0,0) (0,1) | (0,1) (0,2) ... (1,0) (1,1) | (1,1) (1,2) ... (1,0) (1,1) | (1,1) (1,2) ... (2,0) (2,1) | (2,1) (2,2) ... ... .... I would the average of each group of 4 and then get standard deviation and variance for these. In #2 the edges (first and last row, first and last column) and in particular the 4 corner cells would be in fewer calculations. Not a big deal if n is large. The advantage to #2 is that the sample size is larger. Is there anything other to consider that might skew the results one way or the other?