Single indicator variables

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
sgudmundsson
PLS Junior User
Posts: 6
Joined: Mon Dec 19, 2005 12:37 pm
Real name and title:

Single indicator variables

Post by sgudmundsson »

There has been some discussion about single indicator latent variables in the forums. In most article I found the general assumption seems that one should use at least two indicators per LV. Chin, Marcolin and Newsted (1996) showed in a table (no 4) a sample size test with one indicator (significance for the mean estimates). Is this good enough justification for single indicator LV, i.e. you justify it by using larger sample sizes (> 200)? Any feedback on this issue would be highly appreciated.
Top
pipposca
PLS Junior User
Posts: 5
Joined: Tue Nov 07, 2006 2:31 pm
Real name and title:

Post by pipposca »

Hi,
I think that the use of a single indicator in a LV renders the MV like LV.
Top
sgudmundsson
PLS Junior User
Posts: 6
Joined: Mon Dec 19, 2005 12:37 pm
Real name and title:

Post by sgudmundsson »

Thank you for your comment. I think you are absolutely correct on that.

Does anyone know of published PLS studies using one or more single indicator variables? I am also interested to know if the sample size plays any role in the inclusion of such variables in the model?
Top
JKUB
PLS Junior User
Posts: 4
Joined: Tue Jan 30, 2007 12:46 pm
Real name and title:

Single indicator LVs in a PLS model

Post by JKUB »

Hi.

I have used this in a model that was published in Internet Research. I encountered no problems. Pls find the complete reference below.

Brock, J.K.-U. and Zhou, Y. (2005): Organizational Use of the Internet: Scale Development and Validation. In: Internet Research: Electronic Networking Applications and Policy 15(1), pp.: 67-87.
Top
Post Reply

Return to “Method and application”