Use of single items in PLS SEM

[Merin Jacob]

Can I use single item to measure an exogenous variable in my model? Can PLS SEM accomodate single item constructs?

[Geoffrey Hubona]

Yes, I think you can in a limited amount (if it is exogenous). But what happens, of course, is that the so-called "latent" variable to which it is associated is not 'latent' at all. The value of the indicator simply gets "passed along" unchanged.....so you are usurping the purpose of modeling so-called "latent" variables (that cannot be directly measured) altogether.

Further, if it is a single item measure, it makes no different whether it is modeled as a reflective construct or as a formative construct, from a mathematical point of view.....the results will be identical either way......so this also defeats the purpose of using PLS in a sense.

I believe that all of what I say is correct from the mathematical perspective.

Best, Geoffrey Hubona Ph.D.

On Sat, Apr 30, 2016 at 9:22 AM, Merin Jacob <merin...@gmail.com> wrote:

Can I use single item to measure an exogenous variable in my model? Can PLS SEM accomodate single item constructs?

--
Join us at the PLS Applications Symposium: http://plsas.net
---
You received this message because you are subscribed to the Google Groups "PLS-SEM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pls-sem+u...@googlegroups.com.
To post to this group, send email to pls...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Geoffrey Hubona]

So from a measurement and estimation point-of-view, single item exogenous variables can be implemented. However, single item endogenous (predicted) variables will cause problems.

But from a specification point-of-view......that is, what are we trying to accomplish and why are we using path models with latent variables, you get into a "paradigm problem" using any single measurement items, exogenous or endogenous.

They defeat the purpose of using models with latent variables as the variables are no longer latent.

Any one can no longer make a formative vs. reflective distinction using single item measures.

So they are best to avoid their use in PLS path models......use regression instead if you can (or some other alternative modeling approach) or modify your model so they are not needed at all.....

Best, Geoffrey Hubona

[Akwesi Assensoh-Kodua]

Hi Dr. and the good people out there!

 

Thank you  all for sharing. I want to model my discovered factors that are influencing mobile banking in South Africa as a simple bar chart.

Since I do not want to show much details about these factors, I only want to show them as a bar chart with WarpPLS or Excel but have a challenge here.

 

Can the good people out there help me do this, please?

 

Thank you all.

 

Akwesi Assensoh-Kodua.

---

From: pls...@googlegroups.com <pls...@googlegroups.com> on behalf of Geoffrey Hubona <ghu...@gmail.com>
Sent: 30 April 2016 04:13 PM
To: pls...@googlegroups.com
Subject: Re: [pls-sem] Use of single items in PLS SEM

 

---

"This e-mail is subject to our Disclaimer, to view click http://www.dut.ac.za/disclaimer"

Disclaimer

The information contained in this communication from the sender is confidential. It is intended solely for use by the recipient and others authorized to receive it. If you are not the recipient, you are hereby notified that any disclosure, copying, distribution or taking action in relation of the contents of this information is strictly prohibited and may be unlawful.

This email has been scanned for viruses and malware, and automatically archived by Mimecast SA (Pty) Ltd, an innovator in Software as a Service (SaaS) for business. Mimecast Unified Email Management ™ (UEM) offers email continuity, security, archiving and compliance with all current legislation. To find out more, contact Mimecast.

[CMR]

Here is a new article on single items. Enjoy!
http://www.sciencedirect.com/science/article/pii/S014829631630090X

[Merin Jacob]

Dear Prof. Geoffrey Hubona and CMR,

Thanks a lot for your guidance

[Ned Kock]

Merin, single-indicator LVs are okay if the expectation that there is no measurement error is met. Let us consider age, for example. If you measure it as number of years lived after birth, you can assume that there is little to no measurement error involved (other than the month’s approximation). However, if you measure age through “perceptions”, e.g., about how old the respondents feel, then multiple redundant indicators would be advisable to reduce measurement error. If single-indicator perception-based LVs are used in a model, generally the path coefficients will be attenuated. Frequently the power of your test will be reduced. You’ll likely make fewer type I errors as well; i.e., the related hypothesis test or tests will be more conservative. Still, there are advantages in using WarpPLS to analyze models where one, some, or all LVs are measured through single-indicators; one of which is that you’ll be conducting a “robust” path analysis (see links below).

 

How do I conduct a robust path analysis?

 

http://warppls.blogspot.com/2014/03/how-do-i-conduct-robust-path-analysis.html

 

Conducting a nonlinear robust path analysis

 

http://warppls.blogspot.com/2016/02/conducting-nonlinear-robust-path.html

 

Kock, N., & Gaskins, L. (2014). The mediating role of voice and accountability in the relationship between Internet diffusion and government corruption in Latin America and Sub-Saharan Africa. Information Technology for Development, 20(1), 23-43.

 

http://www.scriptwarp.com/warppls/pubs/Kock_Gaskins_2014_ITD_NetCorrup.pdf

 

Kock, N. (2015). Wheat flour versus rice consumption and vascular diseases: Evidence from the China Study II data. Cliodynamics, 6(2), 130–146.

 

http://cits.tamiu.edu/kock/pubs/journals/2015JournalCliodynamics/Kock_2015_Cliodynamics_WheatRiceChina.pdf

 

The links above, as well as other links that may be relevant in this context, are available from:

 

http://warppls.com/

On Saturday, April 30, 2016 at 8:22:45 AM UTC-5, Merin Jacob wrote:

Can I use single item to measure [...]

[Merin Jacob]

Dear Prof. Ned Kock,

Thank you so much or your guidance. 

I have one more query. Can PLS SEM account for reverse causality? It would be really helpful if you could address this query.

On Saturday, April 30, 2016 at 6:52:45 PM UTC+5:30, Merin Jacob wrote: