Standardized coefficients

Home Forums Onyx Helpdesk Standardized coefficients

This topic contains 5 replies, has 4 voices, and was last updated by  admin 1 month ago.

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
  • #416

    Brian Kissell

    Is it possible to obtain standardized coefficients in ONYX?

    Thank you.


    Brian Kissell

    Also, I was curious about whether ONYX is able to provide modification indices.

    Thank you.


    Timo von Oertzen

    Hi Brian,

    Welcome to Onyx! Yes, you can obtain standardized coefficients; right-click on the path with the parameter you are interested in and choose ‘Customize Path’ -> ‘Show Standardized Estimates’. This will give you the standardized estimate on regression paths and the correlation on covariance paths.

    Onyx can give you the modificiation index for any nesting (in particular, for setting a single parameter to zero, which I guess is what you are looking for first); for this, you can copy your model (best with ‘Edit’ -> ‘Clone Model’) and delete the target path there. Then, connect the two models by a path (just right-click on any empty space in one model, drag to the other model, and release the right button there). On the connecting path, there is a dot which will give you the LR (which, for a single parameter, is the MI). I know it would be convenient to get one big table with all MIs for all parameters, it’s on the todo-list, but will take some time (it’s a looong list :).

    Cheers & Happy Eastern,



    Brian Kissell

    Thank you Timo! I am really enjoying the program.



    Hi Timo
    Amos gives you the standardized regression weights even for fixed parameter paths.
    Is there a way I can get those estimates in Onyx too?



    Since version 1.0-1007 (at the time of writing provided as developers’ version), Onyx also shows standardized estimates of fixed paths.

    Thanks for your suggestion!

    • This reply was modified 1 month ago by  admin.
Viewing 6 posts - 1 through 6 (of 6 total)

You must be logged in to reply to this topic.

Posted in