Methods: Data were derived from a self-administered cross-sectional survey study that was mailed to purposive sample of parents who had adopted children from China or Korea between 1991 and 2000 through an international adoption agency. Survey packets were sent to all participants and 1,435 (approximately 81%) completed the packets. The 36-item six-factor structure was tested: Multicultural planning with contact (MPC), multicultural planning with no contact (MPNC), multicultural planning with integration (MPI), Negative attitudes toward racial awareness and survival skills (NRS), Racial awareness (RA), and Survival skills (SS). With the critical value of 78.03 (df=36,1184, p < .05), forty-five observations or approximately 4% of the total valid sample had Mahalanobis distances greater than the critical value. The listwise deletion method was used to manage missing values. Using LISREL version 8.70, with a covariance matrix generated by PRELIS, structural equation model was developed.
Results: CFA resulted in a 36-item six factor structure with acceptable fit: RMSEA (.55 with the 90% confidence interval .046 to .050); NNFI (.96); CFI (.97); IFI (.97); and SRMR (.048). Even though ÷2 showed a lack of fit, the five fit indices – RMSEA, NNFI, CFI, IFI, and SRMR – indicated that the present model fits well to the data. The factor correlations were less than 0.85 and indicated good discriminant validity (Kline, 2005). The factor loadings of NRS were much lower than those of the five other factors (MPC, MPNC, MPI, RA and SS). While three RA items (ra1, ra2, and ra3) loaded at .31, .42, and .51, respectively, ra4 (i.e., It is important for me to remember that others may view my family as different.) loaded at .16, which indicates the variable was not highly related to the latent factor. Most NRS items (nrs1, nrs2, nrs3, nrs4, nrs5, nrs6, nrs8, nrs9, and nrs11) loaded at .26, .26, .21, .22, .22, .29, .24, .29, and .20, respectively.
Conclusions and Implications: Preliminary findings suggest that the TAPS may be useful in examining cultural socialization practices among TRA parents. Consistent with the theory, analyses indicated that cultural socialization practices are a multidimensional construct involving MPC, MPNC, MPI, NRS, RA, and SS. Satisfactory fit indices provided evidence for an acceptable model fit. Negative attitudes toward racial awareness and survival skills were negatively correlated with the other factors. However, double loadings may exist, indicating the factor structure needs to be fine tuned. Future research is needed to further develop TAPS and establish its reliability, validity, and factor structure.