Which lavaan command does JASP 0.15 use to compare models in the SEM module?

Comments

• drpablorogers

  September 2021

  I tried anova and lavTestLRT and couldn't get the same chi-square difference values.
  

  
  

  
      data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABQAAAAUCAQAAAAngNWGAAAA/0lEQVR4AYXNMSiEcRyA4cfmGHQbCZIipkuxnJgMStlMNmeyD2dwmc8+sZgxYJd9ErIZFHUyYYD7fkr6l4/rnvmtl7+KitrqV/fq2Y5eLY3Z9S48eRLe7BmVZ9qhTLhQ0algzZWQOVKSsCF8OjAnwbxDTWFDUhPK/jMr1H6HE/IqRky2DyvCefuwItwZzodVoYRiLqMkVCXrwpJ9twZ+sgfDYEFYl8wIWxZ9uFf7zkallxlJh4YrLGsKjZRx7VGHhLqwgFUN45DGdb8MeXGpgB4ABZdeDcpZEY51A+hyLKz4S1W4MQWm3AibWtgWmk6dyISa1pSdyWTOlLXVp0+eL9D/ZPfBTNanAAAAAElFTkSuQmCC
      
  
  

  
  
• JMBKoch

  September 2021

  Hey drpablorogers,

  
  

  You are using the Santorra Bentler Test here. The basic principle of this test is that it corrects the LR test for certain robust estimation methods (see https://www.statmodel.com/chidiff.shtml). The underlying R-command is lavaan::lavTestLRT(), which automatically applies the Santorra Bentler adjustment method, when the robust estimation methods are applied in making the lavaan object.

  
  

  Best regards,

  Michael
• drpablorogers

  September 2021

  Hello Michael,

  Thanks for your answer. I tried this command (it seems to be the same anova command) but the values ​​don't match with JASP.

  Best regards,

  Pablo Rogers