Chapter 6 182 subgroup, and individual levels, iterations stop if the model fits well (i.e., if two of the four criteria are met: RMSEA ≤ .05, SRMR ≤ .05, CFI ≥ .95, and NNFI ≥ .95). The model will also be pruned, with non-significant associations. Hence, S-GIMME allowed us to identify general (sample-level), subgroup-specific (subgroup-level), and family-specific (individuallevel) associations, all of which had family-specific magnitudes. Figure 2 Summary of model fitting with the S-GIMME algorithm Note. In the current study, same-day (contemporaneous) and next-day (lagged) associations are estimated among 5 parenting practices and 2 adolescent affect network nodes. Notably, the hybrid-GIMME option was used to enable the estimation of data-driven undirected contemporaneous associations that likely exist because of a shared exogenous influence (Luo et al., 2022). These undirected contemporaneous (here sameday) associations reflect the correlations among the variable residuals. They are ideal for
RkJQdWJsaXNoZXIy MTk4NDMw