Skip to main content

Table 1 Goodness of fit statistics for cluster models

From: Clustering of asthma and related comorbidities and their association with maternal health during pregnancy: evidence from an Australian birth cohort

Model LL BIC AIC Npar L2 % reduction in L2 residual df p-value
One class (H0) − 4698.805 9464.953 9415.609 9 1330.987   502 0.000
Two class − 4175.756 8493.682 8389.511 19 284.889 78.6 492 1.000
Three class − 4154.712 8511.456 8363.423 27 242.802 81.8 484 1.000
Four class − 4140.697 8550.771 8353.394 36 214.772 83.9 475 1.000
Five class −4140.679 8625.561 8373.358 46 214.736 83.9 465 1.000
Six class − 4131.795 8637.725 8363.591 50 196.969 85.2 461 1.000
Seven class − 4135.595 8660.289 8375.189 52 204.568 84.6 459 1.000
  1. Abbreviations: AIC Akaike’s Information Criterion, BIC Bayesian Information Criterion, df degrees of freedom, Npar Number of parameters, LL log-likelihood, L2 likelihood-ratio