Chunk #15 — Materials and Methods — Statistical analysis (see Supplementary Methods for details of data preparation) — Multigroup second‐order factor model (CT Study sample)
We hypothesized that the data could be modeled with three first‐order latent factors (CPA, CSA, and PPA) and that the covariance among the first‐order factors can be represented by one second‐order factor, which we term the CTF (see Fig. S1). These analyses were conducted with CFAs appropriate for ordered categorical data (Muthén 1984; Lubke and Muthen 2004) that used mean and variance adjusted weighted least squares estimation (WLSMV). The CFAs, conducted in Mplus 5.2, (Muthen and Muthen 1998) accounted for missing observations. Preliminary CFAs were conducted in female and male groups separately to ensure the second‐order factor model fit the data adequately in both groups. Standard errors were adjusted for nonindependence of observations due to familial clustering. Adequacy of model fit was determined using the comparative fit index (CFI), Tucker‐Lewis index (TLI), and root mean square error of approximation (RMSEA).
