Skip to main content
. 2019 Sep 10;80(4):431–440. doi: 10.15288/jsad.2019.80.431

Table 2.

Fit statistics for Class Solutions 1 through 6

graphic file with name jsad.2019.80.431tbl2.jpg

Variable Number of classes
1 2 3 4 5 6
Fit statistics
 AIC 20,067.378 17,749.241 17,124.424 16,804.148 16,543.830 16,421.349
 BIC 20,137.702 17,859.122 17,273.863 16,993.144 16,772.384 16,689.460
Adjusted BIC 20,086.906 17,779.754 17,165.922 16,856.631 16,607.298 16,495.802
Lo–Mendell–Rubin test – 2,296.242 631.839 332.499 273.565 138.082
 p .024 .522 .666 .320 .856
Entropy – .949 .909 .909 .886 .898

Notes: AIC = Akaike’s Information Criterion; BIC = Bayesian Information Criterion; adjusted BIC = sample size adjusted BIC. Lower values of AIC, BIC, and adjusted BIC indicate a better fitting model. A p value below .05 for the Lo–Mendell–Rubin test indicates that the k class solution fits the data significantly better than the k – 1 class solution. The entropy value ranges from 0 to 1 and provides an indication of classification precision, with values closer to 1 indicating better classification precision.