forum.smartpls.com

Hi everyone,
I am looking for articles on redundancy analysis for second order formative construct using smartpls. I have several reflective first order constructs and this means it is a type II reflective-formative relationship. I have read some articles and book especially by J.M Becker et al 2012 and Hair et al. 2014 but there is no specific explanation on how to analyze that. Thank you very much.

Since 2nd order constructs usually are multidimensional, this kind of analysis does not work for 2nd order constructs.

Best regards
Christian Ringle

cringle wrote:Since 2nd order constructs usually are multidimensional, this kind of analysis does not work for 2nd order constructs.

Best regards
Christian Ringle

Dear Sir,

I am actually having the same question as the OP.
So based on your reply, should we just skip redundancy analysis for the second-order construct?

Best regards
I Putu Ariyasa

A second-order construct itself already is very difficult to establish and to justify content-wise; here it's multidimensional content may consist of the lower order constructs (it is what you define) and no additional tests may be needed.

Best
CR

cringle wrote:A second-order construct itself already is very difficult to establish and to justify content-wise; here it's multidimensional content may consist of the lower order constructs (it is what you define) and no additional tests may be needed.

Best
CR

Dear Sir,

thank you very much for your reply.

I've read from Becker et. al. (2012) that researchers should assess the appropriateness of the higher-order constructs with measurement obtained from the relations between higher-order construct and lower-order constructs. The assessment should match the conceptual properties of the construct (i.e., reflective or formative).
now that based on your previous reply, redundancy analysis is skippable, is there any other assessment that can be skipped too on the formative second-order construct?

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.

Best regards
I Putu Ariyasa

No. But you should really critically ask yourself why you need a 2nd order construct instead of estimating the direct effects from all lower-order constructs to the target construct.

Best
CR

cringle wrote:No. But you should really critically ask yourself why you need a 2nd order construct instead of estimating the direct effects from all lower-order constructs to the target construct.

Best
CR

Thank you very much for your suggestion, i'll do that.

Best regards
I Putu Ariyasa

I am using a HOC full formative. its means I can skip the redundancy analysis test( Convergent validity).
will it possible to have any reference so that I can refer them in my paper