Second Order vs. Interaction Terms

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
Pat79
PLS Junior User
Posts: 9
Joined: Sun Dec 10, 2006 7:19 pm
Real name and title:

Second Order vs. Interaction Terms

Post by Pat79 »

Hello everyone,

this is probably a question for the experts among you, hope there are some in this forum :-).

I just want to clarify when to use a second-order construct and when to use an interaction term approach. To give you more insight a very precise description first, followed by the questions!

I want to test the impact of an organisational concept consisting of 3 subelements: Separation, Integration and Balance. The subelements have further dimensions so I modeled them as First-Order Constructs of which 2 are formative and 1 is reflective. To measure the impact of the organisational concept, all of the 3 elements have to implemented contemporarily.

In my current model, I therefore created a reflective 2-Order Construct for the organisational concept because so its LV depends on the LV of all the 3 subelements. Now I doubt my approach as maybe an interaction term is required to measure the organisational concept and its impact.

So here my questions:

1) What is right approach Second Order Construct or Interaction?
2) If Interaction, how to make an interaction term of 3 constructs that in addition are formulated in different modes (2 reflective, 1 formative).

Need really your help to move on!!! Thanks a lot

Patrick
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,

Used two-stage approach.

Regards,
bwilson
PLS Expert User
Posts: 41
Joined: Wed Apr 02, 2008 7:11 am
Real name and title:
Contact:

Post by bwilson »

Hengkov wrote:Hi,

Used two-stage approach.

Regards,
See The Handbook of Partial Least Squares.
http://link.springer.com/book/10.1007/9 ... 7-8/page/1
Chapter 27. Using PLS to Investigate Interaction Effects Between
Higher Order Branding Constructs.
Best regards,
BradleyWilson
Bradley Wilson. Ph.D.
Senior Lecturer in Advertising.
RMIT University.
School of Media and Communication.
GPO Box 2476V
Location. 9.5.20
Melbourne. Victoria.
Australia.

SEE FOR PUBLICATIONS
www.rmit.edu.au/staff/bradleywilson
Post Reply