Nested and crossed

[rogojel]

hi,
I have the folowing design:

Two paper suppliers, slips of paper and measurement points on each slip. For each supplier we take, say, 3 slips of paper and on each slip we define 6 measurement points. Each slip is painted (same paint, same machine, same operator) and gloss is measured in each point.

Which is the right way to analyze the results?

My first impulse is to consider the slips nested within the factor supplier and the measurement points crossed. On the other hand I could argue that the measurement points are also nested within the slips. Which approach is right?

thanks a lot,
rogojel

 

[Miner]

Slips are definitely nested within supplier. Measurement points are spatial repeated measures by slip.

 

[Dason]

They are spatially repeated but with only six points you probably would be fine ignoring any spatial correlation as long as you're considering them nested within slip (which will automatically add correlation between the measurements on the slips).

 

[rogojel]

Hi,
what we are interested in is whether the positions have a measureable influence on the values e.g. the upper row always has lower values than the lower row, that kind of conclusion. So, analysing the positions as a crossed factor would make sense?

 

[Jake]

In that case, yes, I think position/measurement is a fixed factor crossed with suppliers & papers.