**Youden square** is a reduction from randomized
complete blocks to a balanced incomplete design in which one row or column has been removed from
a Latin square (so it is no longer square). It is a useful design for balancing
out the effects of treatment order in a repeated-measures sequence. The
example on these web pages could pertain to a predator odour treatment (A
with *a* = 4 levels) tainting the food offered to each of *a* = 4 mice
(random factor B). Each mouse is tested with *a*-1 = 3 odour types in order
(random factor C) assigned by the Youden square.

Doncaster, C. P. & Davey, A. J. H. (2007) *Analysis of Variance and Covariance: How to
Choose and Construct Models for the Life Sciences*. Cambridge: Cambridge
University Press.

http://www.southampton.ac.uk/~cpd/anovas/datasets/