Page 1 of 1

composite reliabilities of interaction terms

Posted: Mon Feb 04, 2013 4:18 pm
by spangler
Hi all,

I am testing a model with a number of LVs. The results look good in general. The main effects composite reliabilities are greater than the Cronbach's alphas, but the composite reliabilities of the interaction terms vary and some are much lower than the corresponding alphas. Why is this and what can I do? Thanks.

Cmp Rel Alpha

base sal 1.0000 1.0000
capacity 0.8620 0.7990
implicit agency 0.8853 0.8793
roa 1.0000 1.0000
self 0.8020 0.6927
self * implicit agency 0.8401 0.9400
self * implicit agency 0.3927 0.9400
self * implicit agency 0.5240 0.9400
self * implicit agency 0.7533 0.9400
self * implicit agency 0.0574 0.9400
self * implicit agency 0.5887 0.9400
stck 1.0000 1.0000
stocks 0.9066 0.8458
var 0.8236 0.6790

Posted: Tue Feb 05, 2013 6:22 pm
by Hengkov
Hi,

I not understand why self * implicit agency result six CR and Cronbach Alpha?

My suggestion, before create interaction terms, evaluation outer model and I see your all construct it's OK. Next, create interaction terms and evaluation inner model (no outer model again). Construct interaction terms have low AVE and CR.

Best Regards,
Hengky

Posted: Tue Feb 05, 2013 7:43 pm
by spangler
Hi,

Thanks for the info. I have six dependent LVs in my model, and SmartPLS calculates a separate self-focus x implicit agency interaction for each.

Posted: Wed Feb 06, 2013 10:04 am
by Hengkov
Hi William,
Ok, I understand your research now. ;-)
Good luck

Best Regards,
Hengky

Posted: Thu Feb 07, 2013 2:49 pm
by spangler
Hi Hengky,

Thanks again. Do you know of any published sources that discuss composite reliability and other aspects of PLS interaction terms? It would be nice to have something to cite in our papers submitted for publication.

Posted: Fri Feb 08, 2013 7:41 am
by Hengkov
Hi William,

Check some references interaction below:

Chin, W. W., Marcolin, B. L., and Newsted, P. R. 2003. “A partial least squares latent variable modelling approach for measuring interaction effects: Results from a Monte Carlo simulation study and an electronic-mail emotion/ adoption study,” Information systems research (14:2), pp.189-217.

Dijkstra, T.A., and Henseler, J. 2011. “Linear Indices in Nonlinear Structural Equation Models: Best Fitting Proper Indices and Other Composites,” Quality and Quantity (45), pp. 1505-1518.

Goodhue, D., Lewis, W., and Thompson, R. 2007. “Statistical Power in Analyzing Interaction Effects: Questioning the Advantage of PLS with Product Indicators,” Information Systems Research (18:2), pp. 211-227.

Henseler, J., and Chin, W. W. 2010. “A Comparison of Approaches for the Analysis of Interaction Effects Between Latent Variables Using Partial Least Squares Path Modeling,” Structural Equation Modeling (17:1), pp. 82-109.

Henseler, J and Fassott, G. 2010 “Testing Moderating Effects in PLS Path Models: An Illustration of Available Procedures,” in Handbook of Partial Least Squares: Concepts, Methods and Applications in Marketing and Related Fields, Vincenzo Esposito Vinzi, Wynne W. Chin, Jörg Henseler, and Huiwen Wang, eds., Berlin: Springer, pp. 713-735. (Chapter 30).

Henseler, J., Fassott, G., Dijkstra, T.A., and Wilson, B. 2012. “Analysing quadratic effects of formative constructs by means of variance-based structural equation modeling,” European Journal of Information Systems (21:1), pp. 99-112.

Latan, H., and Ghozali. I. 2012. Partial Least Squares: Concept, Technique and Application SmartPLS 2.0 M3, BP UNDIP. (Chapter 10).

Little, T. D., Bovaird, J. A., and Widaman, K. F. 2006. “On the merits of orthogonalizing powered and product terms: Implications for modeling interactions among latent variables,” Structural Equation Modeling, (13), pp. 497-519.

Martinez-Ruiz, A. 2012. PLS Path Modeling with Mode B. Proceedings: 7th International Conference on Partial Least Squares and Related Methods, Houston, Texas USA.

Best Regards,
Hengky

interaction terms

Posted: Sun Feb 10, 2013 2:04 pm
by spangler
Hi Hengky,

Wow!! This is exactly what we need. Thanks again.