ABSTRACT
Thus far in this book we have said very little about the use of models, either biological or mathematical, for the processes being investigated. Nevertheless, the analysis and interpretation of data that we have discussed in the preceding chapters have been based on implicitly assumed or, in some cases, explicitly stated models. As emphasized in Chapter 8, most of our analyses of experimental data assume that each yield from an experimental unit is the sum of a unit effect and the effect of the treatment applied to that unit. In addition, the set of unit effects is assumed to be approximately normally distributed. When units are grouped into blocks the unit effect is decomposed into a block effect and the residual unit, or error, effect. These assumptions can be written formally as a model,
y
=
b
+
t
+
e
.
