If in the analysis of variance the numbers of observations in the cells defined by the grouping variables are equal then also random effects mixed models are possible. Those grouping variables, whose effects on the dependent variable are regarded as random, are given by the parameter RANDOM, e.g. ANOVA IEADATA,30 DEPENDENT=WEIGHT GROUPING=MICE,FOOD MICE=1(1)4 FOOD=1,2,3 INCLUDED=MF RANDOM=MICE METHOD=TWOWAY where the main effect of MICE is regarded as random and FOOD as fixed. Their interaction is regarded as random. Further information: 1 = More on definitions of the grouping structure 2 = More on analysis of variance for multiple factors A = More on ANOVA operations