The first step of the analysis was to develop a variance components model without predictors.
However, the pseudo R2 was only 7%%, indicating that the unconditional growth model without predictors explains little of the variation in the dependent variable.
First, we tested the latent growth model for cue-induced craving without predictors.
Empty vs. fitted model refers to the model without predictors vs. the model with the individual predictors.
A model without predictors (not shown) was used as the base for calculating how much variance was explained by the inclusion of the predictors.
In performing the hierarchical linear analysis, we first fit the two-level model without predictors (null model) to calculate an intraclass correlation coefficient (ICC).
In the first stage, we estimated the SRM actor, partner and relationship variance without predictors using a random intercepts for actor, partner, dyad and group (Kenny & Livi, 2009).
In order to consider the rate of classification as impressive, the accuracy rate of classification of the model including the predictor variables would need to be a substantial improvement on the classification ability of the model without predictors.
We calculated the intra-class correlation (ICC) (Merlo et al. 2005a) (a measure of the degree of clustering at the country level) and the proportional change in variance (PCV) (Merlo et al. 2005b) for each variable that was added to a respective model using the null model (M0) (a model without predictors) as a reference.
To compute the global CNV significance p-value, the CNVtest function can be used as follows: > CNVtest(mod, "LRT") ----CNV Likelihood Ratio Test---- Chi = 18.75453 (df = 2), pvalue = 8.462633e-05 In this example, a Likelihood Ratio Test (LRT) is computed, comparing a model containing CNV to a model lacking CNV (i.e. a model without predictors or the null model).
