Mode Check Plots Curvature
Posted: Tue Jun 09, 2015 7:55 pmHello,
I have read the forum postings for the Mode Check Plots, and my understanding is that these plots give insight on whether a parameter is identifiable from the data. If the log-lik kernel line is flat in the horizonal direction then the parameter is not identified from the data. Hence, the estimate value maybe questionable.
Is my understanding correct?
When I am interpreting the plots how should one reason about the curvature match or difference of the log-lik kernel and log-post line?
Example in the attached file the comOpenness curves match pretty well. Whereas, Alpha1, eta theta are not good matches.
How should interpret the difference between an upward and downward sloping log-lik kernel line?
For instance zeta vs eta.
Lastly, what do the red dots represent?
Any insight would be greatly appreciated and thank you in advance.
Regards,
Richard
I have read the forum postings for the Mode Check Plots, and my understanding is that these plots give insight on whether a parameter is identifiable from the data. If the log-lik kernel line is flat in the horizonal direction then the parameter is not identified from the data. Hence, the estimate value maybe questionable.
Is my understanding correct?
When I am interpreting the plots how should one reason about the curvature match or difference of the log-lik kernel and log-post line?
Example in the attached file the comOpenness curves match pretty well. Whereas, Alpha1, eta theta are not good matches.
How should interpret the difference between an upward and downward sloping log-lik kernel line?
For instance zeta vs eta.
Lastly, what do the red dots represent?
Any insight would be greatly appreciated and thank you in advance.
Regards,
Richard