Search found 13 matches

by yiqingyu
Sat Dec 16, 2017 8:15 am
Forum: Requested Features
Topic: Mediating effect of a nonlinear relationship
Replies: 2
Views: 15382

Re: Mediating effect of a nonlinear relationship

Oh yes, you are right, just add another quadratic term to Y!
Thank you !
by yiqingyu
Mon Dec 11, 2017 4:04 am
Forum: Requested Features
Topic: Mediating effect of a nonlinear relationship
Replies: 2
Views: 15382

Mediating effect of a nonlinear relationship

Dear PLS developers, Sometimes the curvilinear relationship is mediated by a third variable, like X^2 -->Mediator-->Y. To test the direct effect from the quadratic term X^2 to Y, one has to establish an arrow from X^2 to Y, which is however currently not feasible in SmartPLS. Is it possible to integ...
by yiqingyu
Tue Dec 05, 2017 2:20 am
Forum: Application of PLS
Topic: direction of arrows in the measurement model
Replies: 2
Views: 1897

Re: direction of arrows in the measurement model

Ah! In nature, the mode rather than model specification, determines loadings, weights and path coefficients. Thank you very much for the detailed explanation!
by yiqingyu
Sun Dec 03, 2017 8:41 am
Forum: Application of PLS
Topic: direction of arrows in the measurement model
Replies: 2
Views: 1897

direction of arrows in the measurement model

Hi PLS users, Since the direction of arrows in the measurement model does not indicate whether a factor or composite model is estimated, but Mode A/B instead (Henseler et al. 2016, Sarstedt et al. 2016), 1. why does not the direction of arrows change when I set indicator weighting to a specific Mode...
by yiqingyu
Mon Jul 16, 2012 12:10 pm
Forum: FAQ (SmartPLS 2.0)
Topic: lanten variable score partially reported
Replies: 2
Views: 1497

I checked it again!
That must be due to missing value, i.e. the reported excludes the cases with missing value.
by yiqingyu
Sun Jul 15, 2012 12:00 pm
Forum: FAQ (SmartPLS 2.0)
Topic: lanten variable score partially reported
Replies: 2
Views: 1497

lanten variable score partially reported

Hi all,
I came across sth weird. In the sample there are 453 cases, but when I had a look at the latent variable score (unstandardized), there were only 256 cases reported, where are the score of the other cases?
Thank you!
by yiqingyu
Mon Jul 02, 2012 2:51 pm
Forum: FAQ (Methodology)
Topic: effect size
Replies: 7
Views: 2942

Yes! C functions like a mediation!
SmartPLS doesn´t work for this case?
by yiqingyu
Mon Jul 02, 2012 8:42 am
Forum: FAQ (Methodology)
Topic: effect size
Replies: 7
Views: 2942

Hi,
Thank you for the reply. But I think you got me wrong. There are 3 predictors for D, namely A, B, C. The question is how to calculate effect size for C?
[/img][/code]
by yiqingyu
Sun Jul 01, 2012 9:14 am
Forum: FAQ (Methodology)
Topic: effect size
Replies: 7
Views: 2942

Hi Hengky,
As I have showed, in my model there are 2 exogenous varibles, namely a & b, and I want to calculate the effect size of c...
by yiqingyu
Fri Jun 29, 2012 7:54 am
Forum: FAQ (Methodology)
Topic: effect size
Replies: 7
Views: 2942

effect size

Hi all, I want to calculate effect size f² for c, and my structural model is like: a-->c-->d b-->c-->d b-->d a-->d if I calculate f² for c, i.e. I have to get the value of R² when c is included and excluded respectively, should I delete the whole LV c out of the model, or just delete the path from c...
by yiqingyu
Thu Jun 28, 2012 1:56 pm
Forum: FAQ (Methodology)
Topic: how to choose construct in blindfolding
Replies: 0
Views: 1283

how to choose construct in blindfolding

Hi all
I think it is only meaningful to calculate the Q square of exogenous variables which are predictors as well. Why in the construct option of blindfolding, the endogenous variables are also available? What does it mean if I choose any of the construct to calculate the Q square?
Thank you!
by yiqingyu
Thu Jun 28, 2012 1:52 pm
Forum: FAQ (Methodology)
Topic: Blindfolding: a contradiction
Replies: 0
Views: 1358

Blindfolding: a contradiction

Hi all, I am baffled about the cv communality and cv redundancy, and checked the relevant discussions in the forum, below is the findings: " two latent exogenous variables (LV1 and LV2) and one latent endogenous variable (LV3): for the cv-redundancy of LV3, if you check only the specific endogenous ...
by yiqingyu
Fri Mar 30, 2012 9:34 am
Forum: FAQ (Methodology)
Topic: If LV is measured by ordinal scale
Replies: 1
Views: 980

If LV is measured by ordinal scale

Hi, all
if the dependent LV is simply measured by an ordinal scale, (such as 1=stong, 2=weak, 3= non), is it possible to test it in SmartPLS, and how? Thank you!

Yiqing