The advantage of modelling agreement on a categorical scale among obse
rvers rather than using summarizing indices is now well established. H
owever, analysis of agreement among more than two observers is essenti
ally based on pairs of observers. We present a global and partial agre
ement modelling approach derived from the quasi-independence and quasi
-symmetry log-linear models. This approach addresses high order intera
ctions in the contingency table rather than two-way interaction in the
pairwise agreement approach. Pairwise, global and partial agreement m
odels were applied to the detection by six pathologists of three lesio
ns in biopsy specimens arising from patients suspected to be affected
by Crohn's disease. The global and partial agreement approach surpasse
s the pairwise agreement approach, especially if there is heterogeneit
y among ratings. (C) 1998 John Wiley & Sons, Ltd.