Hello Daniela

first, please refrain from presenting code and output in the text field. Unfortunately, RG hasn't managed it that users can incorporate code....Better is to attach a text file in which you simply copy the code.

BTW: here is an attempt to formulate some rules to present models that help us to help you :)

https://www.researchgate.net/post/Netiquette_proposal_for_questions_on_structural_equation_modeling

I assume that "overall" is meant to represent a second-order factor. In the first model, this factor is not identified.

In addition, you use the anova function on the lavaan object. I never saw this. Are you sure that this works (i tried a small sim and also got an error).

Best,

Holger

Thank you for your answers! I found out that the models (1 and 3; 2 and 3) are not nested, this is were the error message comes from. Dear Holger Steinmetz, thank you for your comment. For model 1, I forgot to add the second order factor "overall". I corrected this mistake above. The anova function (the output is a Chi² difference test) is working with model 1 and 2, which are nested. For future questions, I will follow your guidelines. Thank you and best wishes!