How to settup a databas for PLS model different scenarios- same individuals

Research topics can be discussed in this area.
Post Reply
AnnG
PLS Junior User
Posts: 1
Joined: Sun Jan 10, 2021 5:08 pm
Real name and title: Anna Student Researcher

How to settup a databas for PLS model different scenarios- same individuals

Post by AnnG »

I am currently doing my Master Thesis on how the Willingness to Pay (WTP) and Purchase intention (PI) of customers is affected by the level of customization of a product; I am using PLS-SEM model.

With the survey I made I collected 150 individuals' WTP and PI for two different level of personalization (L1, L2) in 2 different context (C1, C2) and their personal information (age, sex, fashion consciousness.. ecc). I am having some troubles now setting up the model for the PLS analysis because I do not understand how to create separate scenarios (C1L1 and C1L2; C2L1 , C2L2); I do not understand how to create a model where from one individual I obtain different results that have to be interpreted separately, but which are somehow correlated (they belong to the preferences of same customer ). Do you have any suggestion?
jmbecker
SmartPLS Developer
Posts: 1281
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: How to settup a databas for PLS model different scenarios- same individuals

Post by jmbecker »

If I understand you correctly you have an experimental design.
You can simply add your design variables (scenarios) as dummy variables (0,1) and regress them on the outcomes while controlling for other variables (age, sex, fashion consciousness.. etc).
Dr. Jan-Michael Becker, BI Norwegian Business School, SmartPLS Developer
Researchgate: https://www.researchgate.net/profile/Jan_Michael_Becker
GoogleScholar: http://scholar.google.de/citations?user ... AAAJ&hl=de
RMI
PLS Junior User
Posts: 2
Joined: Sat Apr 16, 2022 11:49 am
Real name and title: Romina

Re: How to settup a databas for PLS model different scenarios- same individuals

Post by RMI »

Hello

I have three discounts that I want to examine the effect of them on "perceived deception". In addition, I want to assess these three discounts in two groups of hedonic and utilitarian products. So I have a total of 6 scenarios (normal discount / hedonistic product, normal discount / utilitarian product, tensile discount / utilitarian product, etc.)

My questionnaire structure is based on these 6 scenarios. The system show respondent a hedonic product with a regular discount for example and then ask questions about other variables (perceived deception, perceived risk and etc.)
So as you might notice, I don't have any questions about discount variables on my questionnaire.

Now I'm wondering how can I define discount variables in pls (cause they don't have any indicators)

I searched on the web, and the only answer I got is to use dummy variables (as you have explained here too).
but when I use all 6 dummy variables I've created, all of them reject (t value).
I read that only 2 dummy variables can be used in smart pls. is that right? so should I assess each dummy variable alone or two by two?

I'm not really sure what the right answer to my problem is.
I'd very much appreciate it if you can help me with this.
RMI
PLS Junior User
Posts: 2
Joined: Sat Apr 16, 2022 11:49 am
Real name and title: Romina

Re: How to settup a databas for PLS model different scenarios- same individuals

Post by RMI »

Hello again

I found out that using "dummy variable" is right to do when you have categorical variables like different scenarios.
but the only important point is to avoid the dummy trap (if you have k categories, you can only add k-1 dummy variables to your model)

So in my case, I added 5 dummies to the pls model (I had 6 scenarios) and kept the last one as the reference, and compared the Bootstrap results with the reference dummy variable.

I found some useful explanations in 2 links below:
https://stattrek.com/multiple-regressio ... ables.aspx
https://www.youtube.com/watch?v=YN0I8Dy7Alk

Hope it helps the others.
Post Reply