Can someone tell me about the best values for AVE and GOF?

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
salah
PLS Junior User
Posts: 4
Joined: Fri May 10, 2013 9:22 am
Real name and title:

Can someone tell me about the best values for AVE and GOF?

Post by salah »

Hi there :)

I am new to this software. My sample size is not that large and = 69. I would like to conclude about the significance of my model. Can someone tell me what would be best values for AVE and GOF so my hypohesized model would fit my data?

Please advise

Salah
User avatar
Diogenes
PLS Super-Expert
Posts: 899
Joined: Sat Oct 15, 2005 5:13 pm
Real name and title:
Location: São Paulo - BRAZIL
Contact:

Post by Diogenes »

Hi,

AVE > 0,5
GoF > 0,36

See:
Tenenhaus, M., Esposito Vinzi, V., Chatelin, Y.-M., & Lauro, C. (2005). PLS path modeling. Computational Statistics & Data Analysis, 48(1), 159–205. doi:10.1016/j.csda.2004.03.005

Henseler, J., Ringle, C. M., & Sinkovics, R. R. (2009). The use of partial least squares path modeling in international marketing. Advances in International Marketing, 20, 277–319. doi:10.1108/S1474-7979(2009)0000020014

Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: indeed a silver bullet. The Journal of Marketing Theory and Practice, 19(2), 139–152. doi:10.2753/MTP1069-6679190202

Use the “search” function in this forum (right corner) this question already was answered.

Best regards,

Bido
salah
PLS Junior User
Posts: 4
Joined: Fri May 10, 2013 9:22 am
Real name and title:

Post by salah »

Thank you Prof.

Salah
Emma Wijnen
PLS Junior User
Posts: 10
Joined: Thu Jun 27, 2013 3:00 pm
Real name and title:

GoF values for formative constructs

Post by Emma Wijnen »

Is it true that I cannot calculate the GoF when I have a model that only has formative constructs? Because for the GoF I need the communality (AVE) and those values are not given for a model with formative constructs?

If I am right, what else can I do to assess the goodness of fit?
nayabeyes
PLS Expert User
Posts: 74
Joined: Wed Nov 23, 2011 7:34 am
Real name and title:

Re: GoF values for formative constructs

Post by nayabeyes »

Hi,
Yes you are right.

Assessment of structure model (Hair et al., 2013. A primer on PLS-SEM - Book)
1. Check collinearity among the constructs
2. Bootstrapping to assess the significance of the path coefficient
3. R^2
4. Effect size (f^2)
5. Predictive relevance (q^2)
6. Do not use Gof
MURAD ALI, Ph.D
El Lokalmatador
PLS Junior User
Posts: 2
Joined: Sun Sep 01, 2013 5:31 pm
Real name and title:

Post by El Lokalmatador »

I have a model with a few single-item-constructs. All other constructs have reflective items.

When calculating GoF, shall I leave the singe-item-constructs out? Because of their AVE/Communalities of 1 they wind the GoF up.

Many thanks in advance

*edit:
I found the answer in this paper:

Henseler, J. and Sarstedt, M. (2012), "Goodness-of-fit indices for partial least squares path modeling," Computational Statistics.

As I thought it is recommended to leave singel-item constructs out of GoF-calculation for the reason I stated above.
User avatar
Hengkov
PLS Super-Expert
Posts: 1599
Joined: Sun Apr 24, 2011 10:13 am
Real name and title: Hengky Latan
Location: AMQ, Indonesia
Contact:

Post by Hengkov »

Hi,

This is problem in formula GoF Absolute. If you leave single items construct out, the result will low and not explain models. Contrary, if you enter single items construct, the result will high and misleading. So, for alternative, you must compute GoF Relative.

Regards,
bwilson
PLS Expert User
Posts: 41
Joined: Wed Apr 02, 2008 7:11 am
Real name and title:
Contact:

Post by bwilson »

Hengkov wrote:Hi,

This is problem in formula GoF Absolute. If you leave single items construct out, the result will low and not explain models. Contrary, if you enter single items construct, the result will high and misleading. So, for alternative, you must compute GoF Relative.

Regards,
Hi,
The PLS literature has not had much success in reporting and developing relevant GOF measures to date in my opinion.
Reviewers and journal editors do not look for them..

SO do not use it yet! That would be my suggestion.
regards,
Bradley Wilson. Ph.D.
Senior Lecturer in Advertising.
RMIT University.
School of Media and Communication.
GPO Box 2476V
Location. 9.5.20
Melbourne. Victoria.
Australia.

SEE FOR PUBLICATIONS
www.rmit.edu.au/staff/bradleywilson
bwilson
PLS Expert User
Posts: 41
Joined: Wed Apr 02, 2008 7:11 am
Real name and title:
Contact:

Post by bwilson »

[quote="bwilson"][quote="Hengkov"]

Why not do a power analysis instead given your small sample size.

that would be more useful

brad
Bradley Wilson. Ph.D.
Senior Lecturer in Advertising.
RMIT University.
School of Media and Communication.
GPO Box 2476V
Location. 9.5.20
Melbourne. Victoria.
Australia.

SEE FOR PUBLICATIONS
www.rmit.edu.au/staff/bradleywilson
Post Reply