Table 2. CFA model fit for the VEMI in three samples.
| df | χ2 | CFI | RMSEA | |
|---|---|---|---|---|
| Undergraduate sample 1 | 87 | 359.97 | .975 | .056 |
| MTurk sample 2 | 87 | 462.52 | .975 | .065 |
| LISS sample 3 | 87 | 3813.69 | .948 | .086 |
| Invariance tests across three samples | ||||
| Configural | 261 | 4636.19 | .955 | .080 |
| Constrain factor loadings | 285 | 4804.38 | .954 | .078 |
| Constrain intercepts | 309 | 5978.15 | .942 | .084 |
| Constrain latent means | 315 | 6306.12 | .939 | .086 |
| Invariance tests across males and females in three samples (N = 7,753) | ||||
| Configural | 174 | 3884.0 | .962 | .074 |
| Constrain factor loadings | 186 | 3906.7 | .962 | .072 |
| Constrain intercepts | 198 | 3969.1 | .961 | .070 |
| Constrain latent means | 201 | 4166.3 | .959 | .071 |
| Invariance tests across white and non-white participants in samples 1 and 2 (N = 2,010) | ||||
| Configural | 174 | 767.38 | .977 | .058 |
| Constrain factor loadings | 186 | 784.62 | .977 | .057 |
| Constrain intercepts | 198 | 813.65 | .976 | .056 |
| Constrain latent means | 201 | 822.39 | .976 | .055 |
df = degrees of freedom. CFI = Confirmatory Fit Index. RMSEA = Root Mean Square Error of Approximation. We used model comparison tests based on fit indices to examine measurement invariance. We established measurement invariance (all ΔCFI < .01 or all ΔRMSEA < .01; cf. Cheung & Rensvold, 2002). That is, we were able to constrain configuration, factor loadings, intercepts, and latent means across groups without a significant decrease in model fit.