Table 3.
Fit statistics for the subjective well-being measurement models by country, with chi-square tests of difference
| Models of subjective well-being | χ2 | df | p-value | CFI | TLI | RMSEA |
|---|---|---|---|---|---|---|
| Luxembourg | ||||||
| Model 1: One-factor model | 12.98 | 5 | .02 | .97 | .95 | .06 |
| Model 2: Two-factor model | 6.85 | 4 | .14 | .99 | .98 | .04 |
| Germany | ||||||
| Model 1: One-factor model | 25.46 | 5 | <.001 | .96 | .92 | .10 |
| Model 2: Two-factor model | 12.76 | 4 | .01 | .98 | .96 | .07 |
| Brazil | ||||||
| Model 1: One-factor model | 34.70 | 5 | <.001 | .94 | .87 | .10 |
| Model 2: Two-factor model | 15.58 | 4 | .004 | .98 | .94 | .07 |
| Model 1–Model 2 | Δχ2 | Δdf | p-value |
|---|---|---|---|
| Luxembourg | 6.13 | 1 | .01 |
| Germany | 12.70 | 1 | <.001 |
| Brazil | 19.12 | 1 | <.001 |
Note: Δχ2 = χ2 test of model differences. Model 2 is represented in Figure 1.
CFI = Comparative Fit Index; RMSEA = Root Mean Square Error of Approximation; TLI = Tucker–Lewis Index.