AMOS - ECVI Comparison and Confidence Intervals

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

stat4pat
Posts: 5
Joined: Tue Sep 02, 2008 5:45 pm

AMOS - ECVI Comparison and Confidence Intervals

Postby stat4pat » Tue Sep 02, 2008 5:57 pm

AMOS gives us the ECVI score and the 90% Confidence Intervals for our models. What I am wondering is how we can compare different models using the ECVI. For example, I would like to be able to say "Examination of the ECVI scores show that model X provides significantly better fit to the data than Model Y".

I know that we can use confidence intervals to talk about 'significance' and I have been told that if the confidence intervals don't overlap then that equates to significance. I'm sure that is the case but AMOS only provides 90% confidence intervals for the ECVI, and thus models may still be significantly different even if there is some overlap in their 90% confidence intervals. Is there anyway to change the confidence intervals in AMOS (e.g., to 95% confidence intervals for ECVI) or is there a better way to directly compare ECVI scores and state if they are significantly different or not?

So if anyone has any advice for me I would be very appreciative.

Thank you,

Pat

Who is online

Users browsing this forum: No registered users and 2 guests

cron