Chunk #18 — Materials and Methods — Statistical analysis (see Supplementary Methods for details of data preparation) — Replication analyses (CAT Study sample)
Less preparation was required for these data due to the modifications made to Christchurch Trauma Assessment (Fergusson et al. 1989, 1996a,b; Fergusson and Lynskey 1997) for the project (detailed above). Nonbinary ordinal responses were similarly recoded and some low endorsement items were again combined. Changes reduced the number of items included in the factor analysis to 32 (CPA = 9, CSA = 8, PPA = 15). Multigroup CFAs were then similarly performed as was done for the CT Study data. Due to the streamlined nature of the trauma assessment, combining the CT Study and CAT Study samples in a single multigroup model was not feasible.
