Hodnocení modelu

Structural Equation Modeling with lavaan in R

Erin Buchanan

Professor

Standardizovaná sycení

summary(visual.fit, standardized = TRUE)

Latent Variables:
           Estimate  Std.Err  z-value  P(>|z|)   Std.lv  Std.all
  visual =~                                                             
    x1        1.000                               0.621    0.533
    x2        0.586    0.139    4.215    0.000    0.364    0.310
    x3        0.882    0.149    5.923    0.000    0.548    0.485
    x7        0.728    0.137    5.320    0.000    0.452    0.416
    x8        0.944    0.143    6.599    0.000    0.586    0.580
    x9        1.205    0.170    7.095    0.000    0.748    0.742

Indexy shody

Shoda modelu – jak dobře data odpovídají zadanému modelu
- Indexy dobré shody, např. Comparative Fit Index (CFI) nebo Tucker Lewis Index (TLI)
- Indexy špatné shody, např. Root Mean Square Error of Approximation (RMSEA) nebo Standardized Root Mean Square Residual (SRMR)
Měření shody modelu

Indexy shody

summary(visual.fit, standardized = TRUE, 
        fit.measures = TRUE)

User model versus baseline model:
  Comparative Fit Index (CFI)                    0.701
  Tucker-Lewis Index (TLI)                       0.502
                        _ _ _
Root Mean Square Error of Approximation:
  RMSEA                                          0.190
                        _ _ _
Standardized Root Mean Square Residual:
  SRMR                                           0.111

Pojďme si procvičit!

Structural Equation Modeling with lavaan in R