Dissecting the phenotypic heterogeneity in sensory features in autism spectrum disorder: a factor mixture modelling approach

Table 2 Comparison of different structural models of sensory symptoms in ASD, fit indices and subgroup proportions (N = 332)

Model	Log-likelihood	Par	AIC	BIC	Adjusted BIC	LMR p value	BLRT p value	Class percentages
Latent class analysis
Two-subgroup	− 19,313	115	38,856	39,294	38,929	.004	< .0001	55%, 45%
Three-subgroup	− 18,812	154	37,932	38,518	38,029	.024	< .0001	27%, 33%, 40%
Four-subgroup	− 18,571	193	37,527	38,262	37,650	.141	< .0001	17%, 30%, 15%, 37%
Five-subgroup	− 18,359	232	37,182	38,064	37,329	.160	a	17%, 14%, 34%, 19%, 16%
Six-subgroup	− 18,245	271	37,032	38,063	37,204	.745	a	17%, 10%, 15%, 21%, 22%, 15%
Factor analysis
Five-factor	− 17,623	124	35,495	35,967	35,574	–	–
Six-factor	− 17,563	129	35,384	35,875	35,466	–	–
Seven-factor	− 17,398	135	35,066	35,580	35,152	–	–
Factor-mixture analysis
Two-subgroup, seven-factor	− 17,362	142	35,009	35,549	35,099	.0021	< .0001	19%, 81%
Three-subgroup, seven-factor	− 17,301	150	34,902	35,473	34,997	.01	a	7%, 15%, 77%
Four-subgroup, seven-factor	b

Par number of estimated parameters, AIC Akaike information criterion, BIC Bayesian information criterion, LMR Lo-Mendell-Rubin test, BLRT Bootstrapped likelihood ratio test,
Best performing FMM (‘three-subgroup, seven-factor’) in font bold
^aLog-likelihood was not replicated
^bModel did not converge

ISSN: 2040-2392