GoF and effect size

[kuperbeigo]

I know there are few discussions about GoF including viewtopic.php?t=477 and viewtopic.php?t=841 . Now the only reference that we can make is from Ringle et al (Finite Mixture PLS Analysis) for 'moderate' fit around 0.50. Professor Bido also suggested that above 0.50 can be considered as good.

Now my question is about effect size, since I found it difficult to calculate the effect size. See for example the following:

1. In our study GOF shows to have a value of 0.6184. If we assume a minimum AVE of 0.5 (Fornell and Larcker, 1981) and average medium effect size of 0.09 (Cohen, 1988) GOF would offer us a baseline value of 0.21. (by Van Oppen et al accessed at http://edocs.ub.unimaas.nl/loader/file.asp?id=1066 )

2. Based on the categorization by Cohen (1988), and using 0.50 as cut-off point for communality (Fornell & Larcker 1981), GoF criteria for small, medium and large sizes would be 0.1, 0.25 and 0.36. for the current model, this fit is 0.405, indicating a good fit (by Allard Van Riel accessed at http://books.google.com.au/books?id=zAF ... t&resnum=4 ).

my question is, can anyone show me how did they get those numbers? Thanks in advance.

[Lanfranchi]

Hi,


It is true !

Wetzels, Odekerken-Schröder et van Oppen (2009) suggest the same effects sizes for the same reasons (p.187, end of the page) :

GoF = geometric mean of the average communality and average R².

AVE : cut-off of 0.50 as suggested by Fornell and Larcker (1981)

R² : small : .02, medium : .13, large : .26 proposed by Cohen (1988)

Then

GoF small = .10

GoF medium = .25

GoF large = .36

Cordially yours

Reference

Wetzels, M., Odekerken-Schröder, G. and van Oppen, C. (2009). Using PLS path modeling for assessing hierachical construct models : Guidelines and empirical illustration. MIS Quarterly,33(1), 177-195.

[kuperbeigo]

Thank you A.P. Lanfranchi,

now I got it. Earlier, I did not know how to calculate the effect size. so, is it as the following,

GoF small = (0.50 x 0.02)^0.5 ?
GoF large = (0.50 x 0.26)^0.5 ?

thanks again.

[ruchi]

Hi

If someone has this paper , please provide me, I am not able to download it

Wetzels, M., Odekerken-Schröder, G. and van Oppen, C. (2009). Using PLS path modeling for assessing hierachical construct models : Guidelines and empirical illustration. MIS Quarterly,33(1), 177-195.

Thanks
Ruchita

[alialashwal]

You can download this article from:

http://aureview.com/index.php?option=co ... &Itemid=21