Table 2. Model fit statistics.
| CFI | TLI | RMSEA | # Free | |
|---|---|---|---|---|
| Total sample (N = 43,093) | ||||
| Lifetime diagnoses | .992 | .989 | .012 | -- |
| 12-month diagnoses | .988 | .984 | .010 | -- |
| Women (n = 24,575) | ||||
| Lifetime diagnoses | .993 | .991 | .009 | -- |
| 12-month diagnoses | .990 | .987 | .008 | -- |
| Men (n = 18,518) | ||||
| Lifetime diagnoses | .988 | .984 | .008 | -- |
| 12-month diagnoses | .982 | .976 | .007 | -- |
| Multigroup (Women and Men) | ||||
| Lifetime diagnoses | ||||
| Unconstrained model | .991 | .989 | .012 | 48 |
| Constrained model | .991 | .989 | .012 | 38 |
| 12-month diagnoses | ||||
| Unconstrained model | .987 | .983 | .010 | 48 |
| Constrained model | .988 | .986 | .009 | 38 |
Note: Total sample analyses modeled women and men together. Multigroup analyses modeled women and men simultaneously as two separate groups. Unconstrained models allowed each gender to have unique model parameters; constrained (invariant) models constrained factor loadings and thresholds to equality across genders. CFI: comparative fit index. TLI: Tucker-Lewis index. RMSEA: root mean squared error of approximation. # Free: number of freely estimated parameters.