Help System (web edition)

In the analysis of covariance a part of the independent variables are
at nominal scale (grouping variables),  the others being at interval
scale (covariates). The grouping variables are given by the GROUPING
parameter followed by the definition of their grouping structure.
The covariates are defined by the COVARIATES parameter.

The main effects and interactions of the grouping variables to be
included in the model are defined by the INCLUDED and/or EXCLUDED
parameters. The grouping variables are referenced by their initials.
The default model is always the model with all possible interactions.

A two-way fixed effects analysis of covariance example:

  ANOVA IEADATA,30
  DEPENDENT=KNOWLDGE
  GROUPING=ATTITUDE,SEX
  ATTITUDE=1,2,3  SEX=1,2
  COVARIATES=MENTALPR,APPLICAT

If not otherwise stated the regression coefficients of covariates are
estimated in the total group. If the equality of within-groups regression
coefficients cannot be assumed then the groups, where separate estimates
are wanted, are defined by the parameter SLOPES, e.g. SLOPES=SEX and
SLOPES=ATTITUDE,SEX. The latter would produce different estimates in
the 3 x 2 groups.

Further information:
   1 = More on definitions for grouping variables 
   2 = More on main effects and interactions to be included in the model 
   3 = Methods of forming hypotheses 
   4 = Error terms transferred or assigned by the user 
   E = Estimates of the effects, computing residual and predicted values 
   A = More on ANOVA operations