Skip to main content
. Author manuscript; available in PMC: 2018 Feb 1.
Published in final edited form as: Behav Res Methods. 2017 Feb;49(1):282–293. doi: 10.3758/s13428-015-0697-6

Table 2.

Adjusted Rand index (ARI) results for Simulation I

RV ICEM KM AHC SS
Overall .47 .47 .46 .46 .44
Clusters
 2 .43 .43 .42 .43 .42
 4 .52 .51 .51 .50 .49
 6 .45 .45 .45 .44 .41
Variables
 2 .52 .52 .52 .53 .51
 4 .42 .42 .43 .40 .40
 6 .45 .45 .43 .44 .41
SS
 100 .39 .39 .39 .39 .37
 500 .48 .48 .47 .47 .45
 800 .53 .53 .52 .51 .50
Density
 Equal-sized .44 .42 .44 .44 .43
 1 Small .46 .47 .46 .47 .45
 1 Large .50 .50 .48 .46 .43
Separation
 Close .26 .26 .25 .26 .24
 Moderate .53 .54 .53 .52 .51
 Well .60 .60 .60 .59 .56

Each cell is the average ARI of the classifications of the resulting solution and the true generated cluster structure. Bold values indicate the best ARI for the factor condition. RV, random values; ICEM, iteratively constrained EM; KM, k-means clustering; AHC, agglomerative hierarchical clustering; SS, sum score.