Table 5c.
Evaluation of Invariance Over Time: Dementia (CDR>0.5)
| Model 1 Configural Invariance |
Model 2 Metric Invariance |
Model 3 Strong (Scalar)Factorial Invariance |
Model 4 Strict Invariance |
|||||
|---|---|---|---|---|---|---|---|---|
| Test | Means and variances held constant | Factor Loading/Variances/Means held constant |
Intercepts/Factor Loading/ Variances held constant, means=0 in reference group |
Intercepts/Factor Loading/ Variances / Residual Variances held constant, means=0 in reference group |
||||
| Baseline N=1564 |
Follow-up N=1428 |
Baseline N=1564 |
Follow-up N=1428 |
Baseline N=1564 |
Follow-up N=1428 |
Baseline N=1564 |
Follow-up N=1428 |
|
| Memory | ||||||||
| LM Immediate Recall | 1.04 | 1.06 | 1.03 | 1.07 | 1.03 | 1.06 | 1.04 | 1.04 |
| LM Delayed Recall | 0.63 | 0.59 | 0.61 | 0.61 | 0.62 | 0.61 | 0.61 | 0.61 |
| Attention | ||||||||
| Digits Forward | 0.66 | 0.66 | 0.66 | 0.67 | 0.66 | 0.67 | 0.66 | 0.66 |
| Digits Fwd Length | 0.63 | 0.63 | 0.63 | 0.63 | 0.63 | 0.63 | 0.63 | 0.63 |
| Digits Backward | 0.89 | 0.92 | 0.90 | 0.91 | 0.90 | 0.91 | 0.91 | 0.91 |
| Digits Bkwd Length | 0.85 | 0.88 | 0.86 | 0.88 | 0.86 | 0.88 | 0.87 | 0.87 |
| Executive | ||||||||
| Trails A | 0.85 | 0.85 | 0.85 | 0.86 | 0.85 | 0.86 | 0.85 | 0.85 |
| Trails B | 0.80 | 0.77 | 0.78 | 0.79 | 0.78 | 0.79 | 0.79 | 0.79 |
| WAIS (Digit Symbol) | 0.90 | 0.90 | 0.90 | 0.90 | 0.90 | 0.90 | 0.90 | 0.90 |
| Language | ||||||||
| Animals | 0.90 | 0.90 | 0.89 | 0.90 | 0.89 | 0.91 | 0.90 | 0.90 |
| Vegetables | 0.79 | 0.83 | 0.80 | 0.83 | 0.80 | 0.83 | 0.81 | 0.81 |
| Boston Naming | 0.67 | 0.74 | 0.69 | 0.72 | 0.69 | 0.72 | 0.70 | 0.70 |
| Fit Statistics | ||||||||
| CFI | 0.988 | 0.988 | 0.988 | 0.988 | ||||
| TLI | 0.983 | 0.985 | 0.986 | 0.987 | ||||
| RMSEA | 0.044 | 0.041 | 0.040 | 0.038 | ||||
| SRMR | 0.030 | 0.033 | 0.034 | 0.036 | ||||
| †BIC, p-value, df | −381.56, −, 92 | −464.77, p<.0001, 104 | −522.01, p<.0001, 112 | −601.56, p<.0001, 124 | ||||
| †S-B χ2 p-value, df | df = 92 | p = 0.001, df = 104 | p < 0.01, df = 112 | p = 0.002, df = 124 | ||||
Nested models tested using BIC33 and Satorra-Bentler Scaled χ2test.32 Model 1 was used as the starting point as the least constrained model. Each model was found to be statistically significantly different from the prior model based on both measures
CFI=Comparative fit index, better fit >0.90; range 0.0 -1.0 34
TLI= Tucker-Lewis index: better fit > 0.90; range 0.0 -1.0 36
RMSEA= Root mean square error of approximation: better fit <0.0539
SRMR= Standardized root mean square residual: better fit <0.0540
BIC= Bayesian Information Criterion: better fit<0.033
S-B=Satorra-Bentler Scaled χ2test.32