Hi
I hope anyone can help me with this question since i cannot find a solution on the web (or in this forum).
I have a 2nd order construct type IV (formative-formative) and I am not sure how to model it in smartPLS.
I modelled it as follows:
- 1st order constructs
I modelled the indicators for all three constructs at this level formative, i.e. the arrows point from the indicators to the first order constructs
- 2nd order constructs
I modelled the relation between 1st and 2nd order constructs formative, i.e. the arrows point from the 1st to the 2nd order construct
(now, in my eyes the interesting step)
(a) If I take all 1st order indicators together and drop them on the 2nd order construct and model them formativ, i.e. all arrows point from the indicators to the 2nd order construct, I get really messy results.
(b) If I model the "indicators" (all 1st order items) of the 2nd order construct reflective, I get pretty good results.
Can you please help me and tell me which way, (a) or (b), is the right one.
Thanks a lot, Simon
2nd order constructs type IV (formative-formative)
Re: 2nd order constructs type IV (formative-formative)
Hi Simon,strang wrote: - 1st order constructs
(now, in my eyes the interesting step)
(a) If I take all 1st order indicators together and drop them on the 2nd order construct and model them formativ, i.e. all arrows point from the indicators to the 2nd order construct, I get really messy results.
(b) If I model the "indicators" (all 1st order items) of the 2nd order construct reflective, I get pretty good results.
Can you please help me and tell me which way, (a) or (b), is the right one.
It really depends on how you want to model it... There was a journal of academy of marketing science article on brand relationship quality that tested alternative models like how you specify.
your decision has to be driven by theory AND then considering discriminant validity etc for the formative models.
Many people do not know how to test discriminant validity for formative models. it is an area that needs more work.
See henseler and wilson.. sponsorship papers.
regards
Brad
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
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
Reflective-Formative HOC
Hi,
I have assigned the indicators to the LOC, However I cannot understand how to do it for the HOC,
Please send me a step by step guidelines:)
Cheers,
I have assigned the indicators to the LOC, However I cannot understand how to do it for the HOC,
Please send me a step by step guidelines:)
Cheers,