how to calculate pseudo R2 with mixed ( Two-level model)

[Jennifer Dou]

The syntax:

xtmixed weight week || id:

But i cann't find the Pseudo R2
I calculated it with formula: pseudo R2=di (e(ll_0)-e(ll))/(e(ll_0))
But e(ll_0)=Null???

Please help me and thank you very much

Best,Jennifer Dou

[Carlo Lazzaro]

Jennifer:
welcome to the list.
You may want to take a look at: http://www.statalist.org/forums/foru...ar-mixed-model

[Jennifer Dou]

Thank you very much for the kindly reply, Prof Lazzaro.
I runned the syntax following your guidline, but what a pity, it is the result of ICC but not pseudo R-squared. Since I am a newcomer of STATA and have no ideas on lots of operations, would you mind supplying the detail syntax for pseudo R-squared ?
Thanks soooooo much~

xtmixed Ratee_Per z_Rater_PO M4 || z_Rater_PO:, vce(robust)

Best,
Jennifer

[Joseph Luchman]

Hi Jennifer,

One option for a two-level xtmixed is Snijders & Bosker's (1994) R².

This metric is implemented by the suite mlt (SSC), with the mltrsq module.

- joe

Reference
Snijders, T. A., & Bosker, R. J. (1994). Modeled variance in two-level models. Sociological Methods & Research, 22(3), 342-363.

[Sophie Dibbern]

Does anyone know if there is any command / module to calculate the pseudo R^2 for three-level models? I highly appreciate any advice! Thank you.