See this link: http://lavaan.ugent.be/tutorial/growth.html
### Model
LG.mod <- '
intercept=~1*ALAQSUM+1*CLAQSUM+1*DLAQSUM
slope=~0*ALAQSUM+1*CLAQSUM+2*DLAQSUM
intercept~AAUDITTot.log
slope~AAUDITTot.log
'
fit <- growth(LG.mod, data=A1.tm, missing="ml",group="rndcode")
summary(fit, fit.measures=TRUE, standardized=TRUE)## lavaan (0.5-22) converged normally after 215 iterations
##
## Number of observations per group
## Treatment 362
## Control 763
##
## Number of missing patterns per group
## Treatment 6
## Control 8
##
## Estimator ML
## Minimum Function Test Statistic 2.651
## Degrees of freedom 4
## P-value (Chi-square) 0.618
##
## Chi-square for each group:
##
## Treatment 1.709
## Control 0.942
##
## Model test baseline model:
##
## Minimum Function Test Statistic 754.253
## Degrees of freedom 12
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 1.000
## Tucker-Lewis Index (TLI) 1.005
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -11233.012
## Loglikelihood unrestricted model (H1) -11231.686
##
## Number of free parameters 20
## Akaike (AIC) 22506.024
## Bayesian (BIC) 22606.534
## Sample-size adjusted Bayesian (BIC) 22543.009
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.000
## 90 Percent Confidence Interval 0.000 0.053
## P-value RMSEA <= 0.05 0.938
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.010
##
## Parameter Estimates:
##
## Information Observed
## Standard Errors Standard
##
##
## Group 1 [Treatment]:
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## intercept =~
## ALAQSUM 1.000 7.321 0.882
## CLAQSUM 1.000 7.321 0.782
## DLAQSUM 1.000 7.321 0.878
## slope =~
## ALAQSUM 0.000 0.000 0.000
## CLAQSUM 1.000 3.157 0.337
## DLAQSUM 2.000 6.314 0.757
##
## Regressions:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## intercept ~
## AAUDITTot.log 0.816 0.482 1.693 0.090 0.112 0.102
## slope ~
## AAUDITTot.log -0.668 0.280 -2.385 0.017 -0.212 -0.193
##
## Covariances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .intercept ~~
## .slope -8.998 4.156 -2.165 0.030 -0.399 -0.399
##
## Intercepts:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .ALAQSUM 0.000 0.000 0.000
## .CLAQSUM 0.000 0.000 0.000
## .DLAQSUM 0.000 0.000 0.000
## .intercept 55.775 0.605 92.227 0.000 7.618 7.618
## .slope -1.432 0.335 -4.280 0.000 -0.454 -0.454
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .ALAQSUM 15.241 6.994 2.179 0.029 15.241 0.221
## .CLAQSUM 42.972 4.703 9.136 0.000 42.972 0.490
## .DLAQSUM 13.877 7.581 1.831 0.067 13.877 0.200
## .intercept 53.050 7.945 6.677 0.000 0.990 0.990
## .slope 9.598 3.575 2.685 0.007 0.963 0.963
##
##
## Group 2 [Control]:
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## intercept =~
## ALAQSUM 1.000 7.930 0.937
## CLAQSUM 1.000 7.930 0.799
## DLAQSUM 1.000 7.930 0.907
## slope =~
## ALAQSUM 0.000 0.000 0.000
## CLAQSUM 1.000 4.056 0.409
## DLAQSUM 2.000 8.112 0.928
##
## Regressions:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## intercept ~
## AAUDITTot.log 0.269 0.341 0.788 0.431 0.034 0.031
## slope ~
## AAUDITTot.log 0.107 0.212 0.502 0.615 0.026 0.024
##
## Covariances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .intercept ~~
## .slope -13.730 3.061 -4.486 0.000 -0.427 -0.427
##
## Intercepts:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .ALAQSUM 0.000 0.000 0.000
## .CLAQSUM 0.000 0.000 0.000
## .DLAQSUM 0.000 0.000 0.000
## .intercept 55.282 0.432 128.102 0.000 6.971 6.971
## .slope -2.119 0.256 -8.285 0.000 -0.522 -0.522
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .ALAQSUM 8.813 5.167 1.706 0.088 8.813 0.123
## .CLAQSUM 46.542 3.510 13.259 0.000 46.542 0.473
## .DLAQSUM 2.522 5.269 0.479 0.632 2.522 0.033
## .intercept 62.824 6.069 10.351 0.000 0.999 0.999
## .slope 16.440 2.545 6.459 0.000 0.999 0.999 ##Add Clustering (do not use)
# survey.design <- svydesign(ids=~SCHOOL,prob=~1,data=A1.t,missing="ml")
# survey.fit <- lavaan.survey(lavaan.fit=fit,survey.design = survey.design)
# summary(survey.fit,fit.measures=TRUE,standardized=TRUE,modindices=TRUE)
##Residual correlation
#getCor(fit) #Does not apply to latent growth model