low factor loadings and low AVE and low Cronbachs alpha

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
hxin009
PLS Senior User
Posts: 25
Joined: Mon Aug 13, 2012 4:23 am
Real name and title:

low factor loadings and low AVE and low Cronbachs alpha

Post by hxin009 »

Hi everyone,

I have a construct(reflective) which has three indicators.

The loadings for three indicators are 0.036, 0.817 and 0.819. As you see, one is extremely low and,the AVE for this construct is 0.44, the composite reliability 0.62 and the cronbachs alpha is 0.41. All of them do not look good. So i have to drop one indicator, right?

After i dropped one indicator, the loadings for the rest two indicators are 0.828 and 0.827, the AVE for this construct rise to 0.67, the composite reliability rise to 0.8 and the cronbachs alpha is 0.51. They look much better.

so my question is, shall i drop this indicator with low factor loadings?

if i drop this indicator, then it leaves only two indicators to the construct. I recall it is suggested a construct should at least has three indicators.

Any ideas? Any literature you may suggest?

Thank you very much!


Best regards,
Catherine
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 Catherine,

Yes, Right.
In PLS-SEM two indicator no problem, but your model not consistent at large. In CB-SEM your model unidentified or underidentification if only used two indicator.

Best Regards,
Hengky
hxin009
PLS Senior User
Posts: 25
Joined: Mon Aug 13, 2012 4:23 am
Real name and title:

Post by hxin009 »

Hi Hengky, thanks for your help. Are there any literature i could use for reference? Thank you again! You've been so helpful!


Best regards,
Catherine
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,

See
Tenenhaus, M., EspositoVinzi,V., Chatelin,Y.-M., and Lauro, C. 2005. “PLS path modeling,” Computational Statistics and Data Analysis, (48:1), pp. 159-205. (ECSI Model)

Bobow-Thies, S and Albers, S. 2010. “Application of PLS in Marketing: Content Strategies on the Internet ,” In V.E. Vinzi, W. W. Chin, J. Henseler, & H. Wang (Eds.), Handbook of partial least squares: Concepts, methods and applications in marketing and related fields (pp.589-604).Berlin: Springer.

Best Regards,
Hengky
hxin009
PLS Senior User
Posts: 25
Joined: Mon Aug 13, 2012 4:23 am
Real name and title:

Post by hxin009 »

Hi Hengky,

Thanks you so much! very helpful!

Best regards,
Catherine
Post Reply