High Composite Reliability Query

Frequently asked questions about PLS path modeling.
Post Reply
Sussmek
PLS Junior User
Posts: 8
Joined: Sun Feb 22, 2015 3:28 pm
Real name and title: Sussmek S

High Composite Reliability Query

Post by Sussmek » Mon Mar 02, 2015 9:32 am

Hi,

In my Structural model 2 of the 6 predictor variables (All variables have 4 measurements items each. ) have shown composite reliability of .953 and .967 . Im aware that composite reliability of over .95 is not desirable. Any suggestions what can be if its over? Should i drop these variables? Is it ok to proceed with the analysis?

(All variables have 4 measurements items each. )

Thanks

Sussmek

amoghta
PLS Junior User
Posts: 5
Joined: Sun Mar 08, 2015 5:37 pm
Real name and title: Amogh Talan, Assistant Professor

Re: High Composite Reliability Query

Post by amoghta » Sun Mar 29, 2015 7:58 pm

Check the outer loadings of all the indicators involved in the conflicting constructs. If the outer loading is < 0.40 remove such indicators. If the loading is between 0.4 and 0.7, remove such indicators one by one and see if it improves the composite reliability. If it does, then remove them.

agalvez
PLS Expert User
Posts: 39
Joined: Mon Jul 04, 2016 10:17 pm
Real name and title: Alex

Re: High Composite Reliability Query

Post by agalvez » Sun Sep 25, 2016 4:43 pm

Sussmek wrote:Hi,

In my Structural model 2 of the 6 predictor variables (All variables have 4 measurements items each. ) have shown composite reliability of .953 and .967 . Im aware that composite reliability of over .95 is not desirable. Any suggestions what can be if its over? Should i drop these variables? Is it ok to proceed with the analysis?

(All variables have 4 measurements items each. )

Thanks

Sussmek
Dear Sussmek

Did you solve this?

Regards

Post Reply