Table 2.
Dimensionality models tested using EFA and CFA of the The Coparenting Relationship Scale [22] in the adaptation process to the Spanish population of engaged parents and separated or divorced parents, CRS-SEg-S&D.
| MODEL | χ2 (df) | χ2/df | BIC/ECVI | CFI [TLI] | RMSEA [90%CI] | SRMR | 1 S | |||
|---|---|---|---|---|---|---|---|---|---|---|
| T CFA | CRS6S (K = 30) | 2972.158 (390) | 7.62 | 1312 | 0.968 [0.964] | 0.052 [0.050–0.054] | 0.071 | 0.195 | ||
| C EFA | CRS6S (K = 30) | −123.124 | [0.881] | 0.069 [0.066–0.071] | 0.212 | |||||
| C EFA | M1 (K = 21) | −0.536 | [0.912] | 0.071 [0.067–0.074] | 0.999 | |||||
| V CFA | M1 (K = 21) | 203.218 (188) H | 1.08 | 0.328 | 0.999 [0.999] | 0.008 [0–0.016] | 0.035 | |||
| V CFA | M2 (K = 20) | 176.660 (169) J | 1.04 | 0.226 | 1 [1] | 0.006 [0–0.015] | 0.034 | |||
| V,3 Invariance | M2 Sex | χ2 (df) | χ2/df | CFI | RMSEA [90%CI] | SRMR | ∆CFI | ∆RMSEA | ∆SRMR | |
| Conf. Invar. | 227.651 (338) | 0.673 | 1 | 0 [0–0] | 0.040 | |||||
| Metr. Invar. | 282.733 (356) | 0.794 | 1 | 0 [0–0] | 0.044 | 0 | 0 | 0.004 | ||
| Scal. Invar. | 305.615 (374) | 0.817 | 1 | 0 [0–0] | 0.043 | 0 | 0 | −0.001 | ||
| Strict Invar. | 317.394 (394) | 0.805 | 1 | 0 [0–0] | 0.044 | 0 | 0 | 0.001 | ||
| Invariance | M2 MS | |||||||||
| Conf. Invar. | 300.160 (338) | 0.888 | 1 | 0 [0–0.005] | 0.041 | |||||
| Metr. Invar. | 432.841 (356) | 1.215 | 0.996 | 0.019 [0.012–0.025] | 0.047 | 0.004 | 0.001 | 0.006 | ||
| Scal. Invar. | 505.343 (374) | 1.351 | 0.994 | 0.024 [0.019–0.030] | 0.047 | 0.002 | 0.005 | 0 | ||
| Strict Invar. | 594.265 (394) | 1.508 | 0.991 | 0.029 [0.024–0.034] | 0.050 | 0.003 | 0.005 | 0.003 |
Legend. χ2/df ratio [77]; BIC/ECVI = parsimony indices, BIC information criteria in EFA/expected cross-validation index in CFA; CFI [TLI] = comparative fit index [Tucker–Lewis index]; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Mean Square of Residuals in CFA; S = Bentler’s simplicity index; 1 S was obtained through FACTOR (JASP does not provide the value); CRS6S (K = 30), model found by Feinberg et al. [26], deleting the Exposure to Conflict scale. K = 30 items and 6 factors; T = total sample, N = 2427 participants; C = calibration sample (50% approx., n = 1239); V = validation sample (50% approx., N = 1188); M1 = best fitted model in the EFA (k = 21), where nine items were eliminated; M2 = the CFA of Model M1 fits well. However, it was considered necessary to eliminate item 16 (see the explanation in Table 1). 3 configural, metric, scalar, and strict invariance, respectively; ∆ comparison of the increment of the observed value in CFI, SRMR and RMSEA; MS = marital status (Eg and S&D); H p = 0.212; J p = 0.328. The best-fitting models of the EFA and CFA are highlighted in bold. For the rest, see Table 1.