Table 4.

Number and percentage of 762 dogs presenting with separation related problems assigned to each group according to the type of cluster analysis.

HACA	A (n = 133)	%	B (n = 221)	%	C (n = 271)	%	D (n = 137)	%
CAPAM
E (n = 118)	114	85.714	3	1.358	0	0	1	0.73
%	96.610		2.542		0		0.848
F (n = 244)	17	12.782	200	90.498	6	2.214	21	15.329
%	6.967		81.967		2.459		8.607
G (n = 118)	3	2.256	0	0	13	4.797	104	75.912
%	2.542		0		11.017		88.136
H (n = 282)	1	0.752	18	8.145	252	92.989	11	8.029
%	0.355		6.383		89.362		3.901

Bold and italic numbers indicate to which group most dogs were assigned. Underlined numbers on the right of the bold and italic ones are related to the Hierarchical agglomerative cluster analysis (HACA) while those under the bold and italic are related to the Cluster analysis using partitioning around medoids method (CAPAM).

HACA, hierarchical agglomerative (clusters named A, B, C, and D); and CAPAM, partitioning around medoids (clusters named E, F, G, and H).