Robust Problem - CFA with MLR - Thesis Help

210 views
Skip to first unread message

yalin...@gmail.com

unread,
Jun 12, 2018, 11:15:00 AM6/12/18
to lavaan
Hello, I'm currently dealing with my thesis analysis. I just learned how to use R and I'm inexperienced, so I will be grateful if you could help me.


I have a question regarding the output of confirmatory factor analysis conducted with MLR estimator.

First I'm not sure if I'm doing it right. First could you check my sytax? Am I doing it right?I'm not sure about second order CFA syntax. (Please see attached figure and xls data file.)

Secondly, when I run this syntax,  robust values were not calculated for RMSEA,CFI,TLI...etc. Robust estimates in the table are 'NA'. What should I do? Should I give non-robust estimates?

I don't know what to do, please help me.

Thank you in advance.

Eren Yalın

The syntax I used was like this:

#First order CFA
Model<- 'Fact1=~c3+c6+c8+c12+c14+c17+c24+c25
Fact2=~c5+c7+c10+c20+c22+c23+c29+c31
Fact3=~c1+c2+c11+c13+c18+c19+c21+c30
Fact4=~c4+c15+c16+c26+c27+c32
Fact5=~c9+c28+c33'

fit <- cfa(Model, data=veri, estimator = "MLR", se = "robust.mlr")

summary(fit, fit.measures=TRUE)

#Second order CFA
Model1 <- ' Fact1=~c3+c6+c8+c12+c14+c17+c24+c25
Fact2=~c5+c7+c10+c20+c22+c23+c29+c31
Fact3=~c1+c2+c11+c13+c18+c19+c21+c30
Fact4=~c4+c15+c16+c26+c27+c32
Fact5=~c9+c28+c33
Sleep =~ Fact1 + Fact2 + Fact3+ Fact4 + Fact5'

fit1 <- cfa(Model1, data=veri, estimator = "MLR", se = "robust.mlr")
summary(fit1, fit.measures=TRUE)

RESULTS WERE:

lavaan (0.6-1) converged normally after  64 iterations

 

  Number of observations                           335

 

  Estimator                                         ML      Robust

  Model Fit Test Statistic                    1260.164    1065.183

  Degrees of freedom                               485         485

  P-value (Chi-square)                           0.000       0.000

  Scaling correction factor                                  1.183

    for the Yuan-Bentler correction (Mplus variant)

 

Model test baseline model:

 

  Minimum Function Test Statistic             5269.717    4303.054

  Degrees of freedom                               528         528

  P-value                                        0.000       0.000

 

User model versus baseline model:

 

  Comparative Fit Index (CFI)                    0.837       0.846

  Tucker-Lewis Index (TLI)                       0.822       0.833

 

  Robust Comparative Fit Index (CFI)                            NA

  Robust Tucker-Lewis Index (TLI)                               NA

 

Loglikelihood and Information Criteria:

 

  Loglikelihood user model (H0)             -14642.995  -14642.995

  Loglikelihood unrestricted model (H1)     -14012.913  -14012.913

 

  Number of free parameters                         76          76

  Akaike (AIC)                               29437.990   29437.990

  Bayesian (BIC)                             29727.864   29727.864

  Sample-size adjusted Bayesian (BIC)        29486.784   29486.784

 

Root Mean Square Error of Approximation:

 

  RMSEA                                          0.069       0.060

  90 Percent Confidence Interval          0.064  0.074       0.055  0.064

  P-value RMSEA <= 0.05                          0.000       0.000

 

  Robust RMSEA                                                  NA

  90 Percent Confidence Interval                                NA     NA

 

Standardized Root Mean Square Residual:

 

  SRMR                                           0.069       0.069

 

Parameter Estimates:

 

  Information                                 Observed

  Observed information based on                Hessian

  Standard Errors                   Robust.huber.white

 

Latent Variables:

                   Estimate  Std.Err  z-value  P(>|z|)

  Fact1 =~                                           

    c3                1.000                          

    c6                1.618    0.376    4.297    0.000

    c8                1.600    0.359    4.458    0.000

    c12               1.972    0.487    4.046    0.000

    c14               1.778    0.461    3.859    0.000

    c17               1.940    0.450    4.307    0.000

    c24               1.389    0.338    4.115    0.000

    c25               1.680    0.433    3.881    0.000

  Fact2 =~                                            

    c5                1.000                          

    c7                1.348    0.121   11.150    0.000

    c10               1.298    0.130    9.949    0.000

    c20               0.886    0.115    7.732    0.000

    c22               1.392    0.137   10.174    0.000

    c23               0.796    0.108    7.372    0.000

    c29               1.104    0.106   10.440    0.000

    c31               0.812    0.092    8.787    0.000

  Fact3 =~                                            

    c1                1.000                          

    c2                1.002    0.058   17.338    0.000

    c11               1.047    0.091   11.566    0.000

    c13               1.242    0.090   13.817    0.000

    c18               0.849    0.084   10.050    0.000

    c19               0.751    0.091    8.293    0.000

    c21               0.922    0.103    8.916    0.000

    c30               1.170    0.100   11.720    0.000

  Fact4 =~                                            

    c4                1.000                          

    c15               1.255    0.128    9.794    0.000

    c16               0.828    0.114    7.265    0.000

    c26               1.133    0.103   11.005    0.000

    c27               1.302    0.141    9.226    0.000

    c32               0.574    0.090    6.356    0.000

  Fact5 =~                                           

    c9                1.000                          

    c28               1.226    0.190    6.454    0.000

    c33               1.048    0.179    5.853    0.000

 

Covariances:

                   Estimate  Std.Err  z-value  P(>|z|)

  Fact1 ~~                                           

    Fact2             0.118    0.032    3.623    0.000

    Fact3             0.068    0.028    2.401    0.016

    Fact4             0.089    0.032    2.817    0.005

    Fact5            -0.026    0.016   -1.629    0.103

  Fact2 ~~                                           

    Fact3             0.316    0.052    6.061    0.000

    Fact4             0.328    0.057    5.798    0.000

    Fact5            -0.164    0.043   -3.843    0.000

  Fact3 ~~                                           

    Fact4             0.496    0.070    7.132    0.000

    Fact5            -0.135    0.056   -2.418    0.016

  Fact4 ~~                                           

    Fact5            -0.208    0.049   -4.211    0.000

 

Variances:

                   Estimate  Std.Err  z-value  P(>|z|)

   .c3                0.746    0.084    8.867    0.000

   .c6                0.476    0.068    7.006    0.000

   .c8                0.618    0.099    6.265    0.000

   .c12               0.362    0.079    4.582    0.000

   .c14               0.373    0.070    5.321    0.000

   .c17               0.223    0.037    6.019    0.000

   .c24               0.287    0.047    6.134    0.000

   .c25               0.245    0.044    5.593    0.000

   .c5                1.057    0.086   12.298    0.000

   .c7                0.550    0.070    7.848    0.000

   .c10               0.695    0.078    8.877    0.000

   .c20               1.183    0.092   12.915    0.000

   .c22               0.353    0.052    6.781    0.000

   .c23               0.445    0.054    8.236    0.000

   .c29               0.554    0.069    7.995    0.000

   .c31               0.449    0.073    6.143    0.000

   .c1                0.781    0.079    9.911    0.000

   .c2                0.913    0.090   10.112    0.000

   .c11               0.627    0.068    9.189    0.000

   .c13               0.575    0.083    6.895    0.000

   .c18               0.680    0.063   10.776    0.000

   .c19               1.024    0.082   12.552    0.000

   .c21               1.044    0.090   11.621    0.000

   .c30               0.832    0.080   10.388    0.000

   .c4                1.059    0.094   11.218    0.000

   .c15               0.909    0.102    8.933    0.000

   .c16               0.809    0.078   10.340    0.000

   .c26               0.668    0.067    9.910    0.000

   .c27               0.874    0.105    8.328    0.000

   .c32               0.762    0.066   11.595    0.000

   .c9                1.215    0.157    7.728    0.000

   .c28               1.177    0.204    5.756    0.000

   .c33               1.330    0.175    7.614    0.000

    Fact1             0.120    0.046    2.601    0.009

    Fact2             0.417    0.077    5.442    0.000

    Fact3             0.674    0.102    6.584    0.000

    Fact4             0.537    0.093    5.806    0.000

    Fact5             0.665    0.159    4.189    0.000

SEM.jpg
veri.xlsx

Terrence Jorgensen

unread,
Jun 13, 2018, 9:57:59 AM6/13/18
to lavaan
First could you check my sytax? Am I doing it right?I'm not sure about second order CFA syntax.

Model syntax looks fine.  But when you request estimator = "MLR", that already sets the se= and test= arguments to be robust, so you don't need the se= argument.

Secondly, when I run this syntax,  robust values were not calculated for RMSEA,CFI,TLI...etc. Robust estimates in the table are 'NA'. 

That was a bug that was recently fixed:


You can install the development version, and they should be available.

install.packages("lavaan", repos = "http://www.da.ugent.be", type = "source")

Terrence D. Jorgensen
Postdoctoral Researcher, Methods and Statistics
Research Institute for Child Development and Education, the University of Amsterdam

Reply all
Reply to author
Forward
0 new messages