Skip to main content
. Author manuscript; available in PMC: 2010 Sep 1.
Published in final edited form as: Psychol Sci. 2009 Aug 14;20(9):1146–1152. doi: 10.1111/j.1467-9280.2009.02425.x

TABLE 2. Statistical Comparison of Models 1 Through 6.

Model Description −2LL k LR Δdf p AIC
Model 1 Saturated model 15,097.344 28
Model 2 Means equal for Twins 1 and 2 15,101.134 24 3.790 4 .435 −4.210
Model 3 Means equal for monozygotic and dizygotic twins 15,099.980 24 2.636 4 .620 −5.364
Model 4 Means equal for induction and follow-up 15,171.044 24 73.700 4 <.001 65.700
Model 5 Variances and covariances equal for monozygotic and dizygotic twins 15,340.484 18 243.141 10 <.001 223.141
Model 6 Bivariate ACE model, means differ only between baseline and follow-up 15,122.444 11 25.100 17 .092 −8.900

Note. The table presents the following values: −2 times the log likelihood (−2LL), the number of free parameters in the model (k), the results of a likelihood ratio test (LR, distributed as a chi-square statistic), the change in degrees of freedom of the saturated model and submodel (Δdf), the significance of the chi-square test (p), and an index of the balance between goodness of model fit and parsimony (Akaike's information criterion, AIC). The bivariate ACE model includes additive genetic factors (A), shared family environment (C) and nonshared environment (E).