Goodness -
of -
fit indices for the six - factor model with 14 items
indicate a well - adjusted
fit to the data (χ2 / df = 1.427, RMSEA = 0.019, SRMR = 0.021, CFI = 0.992, AGFI = 0.982) which confirms study 1's findings.
Fit indices used to evaluate the model included a χ2 goodness - of - fit test (nonsignificant values indicate good fits), the comparative fit index (scores of > 0.95 indicate better fits), the root mean square error of approximation (values of < 0.05 indicate good fits), and the standardized root mean square residual (values of < 0.08 indicate good fits).43, 44 Missing values were imputed through multiple imputation by using functions in the missing data library in S - Plus (Insightful Corp, Seattle, WA).45, 46 The combined data for the cross - lagged / survival model converged more quickly with 15 imputed data sets than did the model that used a likelihood - based approach to missing da
Fit indices used to evaluate the model included a χ2
goodness -
of -
fit test (nonsignificant values indicate good fits), the comparative fit index (scores of > 0.95 indicate better fits), the root mean square error of approximation (values of < 0.05 indicate good fits), and the standardized root mean square residual (values of < 0.08 indicate good fits).43, 44 Missing values were imputed through multiple imputation by using functions in the missing data library in S - Plus (Insightful Corp, Seattle, WA).45, 46 The combined data for the cross - lagged / survival model converged more quickly with 15 imputed data sets than did the model that used a likelihood - based approach to missing da
fit test (nonsignificant values
indicate good
fits), the comparative
fit index (scores of > 0.95 indicate better fits), the root mean square error of approximation (values of < 0.05 indicate good fits), and the standardized root mean square residual (values of < 0.08 indicate good fits).43, 44 Missing values were imputed through multiple imputation by using functions in the missing data library in S - Plus (Insightful Corp, Seattle, WA).45, 46 The combined data for the cross - lagged / survival model converged more quickly with 15 imputed data sets than did the model that used a likelihood - based approach to missing da
fit index (scores
of > 0.95
indicate better
fits), the root mean square error
of approximation (values
of < 0.05
indicate good
fits), and the standardized root mean square residual (values
of < 0.08
indicate good
fits).43, 44 Missing values were imputed through multiple imputation by using functions in the missing data library in S - Plus (Insightful Corp, Seattle, WA).45, 46 The combined data for the cross - lagged / survival model converged more quickly with 15 imputed data sets than did the model that used a likelihood - based approach to missing data.
The logistic regression model's χ2 and Hosmer and Lemeshow
goodness -
of -
fit test statistics were respectively significant (χ2 = 31.187, p =.008) and not significant (χ2 = 4.384, p =.821), both
indicating a well -
fitting model; the Nagelkerke's R 2 was moderate (0.459).