For establishing convergent validity 'Redundancy Analysis' is to be carried out in case of formative construct in PLS. It requires use of an existing formative latent variable as an exogenous latent variable to predict an endogenous latent variable operationalized through one or more reflectively measured indicators or a global indicator.
We need to include the global item for a given latent construct in the questionnaire for collection of data. How is the global item/indicator for the same construct created?
If you are at survey developing stage then you need to include a direct question that covers your latent construct as whole. For instance, if one use to check the customer satisfaction about the product, one might ask the question regarding price, quality etc. However, if variable are formative and establish in literature then one might use one direct question about that particular product and take that as global measure. or Second way around is one might make a index for that construct (if you don't include the global measure in survey stage) and proceed further empirical tests and third way for formative model evaluation involve establishing Nomological validity, weights significance and multicollinearity assessment (Henseler et al., 2009). In the end Formative construct assessment is always tricky as formative indicators are not highly correlated together and second, formative indicators are assumed to be error free .Therefore, traditional reflective indicators assessment includes internal consistency and reliability or discriminant
validity becomes ineffective. So one should relay the above stated options.
For details please read
Hair Jr, J. F., Hult, G. T. M., Ringle, C., & Sarstedt, M. (2016). A primer on partial least
squares structural equation modeling (PLS-SEM): Sage Publications.
Hope it would be helpful for you.
This also might be useful : https://www.researchgate.net/publication/261178344_Partial_Least_Squares_PLS_Structural_Equation_Modeling_SEM_for_Building_and_Testing_Behavioral_Causal_Theory_When_to_Choose_It_and_How_to_Use_It
Article Partial Least Squares (PLS) Structural Equation Modeling (SE...
@Rashid Thank you for the answer.
You have suggested to create an index for the construct in case the global measure is not included at the time of data collection. Could you please provide the reference for this.I wish to confirm it as Hair et al (2014) has clearly given "to execute this approach, the reflective latent variable must be specified in the research design phase and included in data collection for the research".
Regards
Sarita
Hi @Rashid, i have a question.
What if the construct is HOC and measured by using reflective-formative indicators,. could we use the reflective indicators of the first-order constructs in the redundancy analysis of second-order formative construct?
Hi @Paulo Duarte can u explain the logic behind? I am also curious for it
Dear Scholars,
I read a feedback from Prof. Dr. Christian M. Ringle in the https://forum.smartpls.com/viewtopic.php?t=15579
and included an article from Becker, Jan-Michael, Kristina Klein, and Martin Wetzels. "Hierarchical latent variable models in PLS-SEM: guidelines for using reflective-formative type models." Long Range Planning 45.5 (2012): 359-394. that redundancy analysis is not required for second order formative construct
Could anyone be so kind to assist to share any current articles that supports the above statement..
Thank you
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity