Balance: A balanced design has the same number of replicate observations in each sample. Balance is a desirable attribute particularly of cross factored models, where loss of balance generally (though not inevitably) leads to loss of orthogonality. The consequent complications to the partitioning of sources of variance in the response are accommodated by General Linear Models.

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/