Table 3.

Statistics to determine the optimal number of consumer segments.

Number of Segments	Number of Parameters (p)	Log Likelihood at Convergence (LL)	AIC a	AIC3 b	BIC c	${\overline{\mathsf{\rho }}}^2$ d
2	8	−434.69	885.37	893.37	446.30	0.05
3	12	−404.53	833.06	845.06	421.95	0.10
4	16	−395.36	822.72	838.72	418.59	0.11

Note: Log likelihood evaluated at zero is −464.20. ^a AIC (Akaike Information Criterion) is calculated using −2(LL − p); ^b AIC3 (Bozdogan Akaike Information Criterion) is calculated using (−2LL + 3p); ^c BIC (Bayesian Information Criterion) is calculated using (−LL + (p/2)xln(N)); ^d ${\overline{\mathsf{\rho }}}^2$ is calculated using [1 − AIC/2LL(0)].