. Author manuscript; available in PMC: 2018 Jun 1.

Published in final edited form as: Multivariate Behav Res. 2012 Jun 18;47(3):463–492. doi: 10.1080/00273171.2012.673952

Table 3. Mean Differences of Adjusted Rand Indices Between PCAPP and Other Methods.

Comparison	Difference in Means	SD	Effect Sizes^†
PCAPP vs. KM	(.51-.26)=.25^**	.29	1.28
PCAPP vs. TA	(.51-.18)=.33^**	.28	1.69
PCAPP vs. RKM	(.51-.31)=.20^**	.30	0.95
PCAPP vs. FKM	(.51-.17)=.34^**	.30	1.65
PCAPP vs. PPK	(.51-.06)=.45^**	.22	3.09

Note. df = 431 for each t-test.

^**

p ≤ .0001, two-tailed.

^†

the effect size was computed assuming independent groups to protect against over-inflating the estimate as recommended by Dunlop, Cortina, Vaslow, and Burke (1996).

KM = K-means clustering; TA = Tandem Analysis; RKM = Reduced K-means; FKM = Factorial K-means; PPK = Projection Pursuit Kurtosis; PCAPP = Principal Cluster Axes Projection Pursuit.