Gender a moderator on SmartPLS 3

Questions about the implementation and application of the PLS-SEM method, that are not related to the usage of the SmartPLS software.
Post Reply
Priyankafox
PLS Senior User
Posts: 21
Joined: Fri Sep 18, 2015 6:40 pm
Real name and title: Ms Priyanka

Gender a moderator on SmartPLS 3

Post by Priyankafox »

Hi all,

How does one make a gender a moderator in a direct relationship?

For instance in the diagram attached.

How would we make gender a moderator between air travel risk perception and mas brand image?

I have tried using the PLS textbook but it mostly guides for SmartPLS 2.0. I would appreciate some key steps on how to do moderation when the variable is a demographic variable.

Many Thanks,
Priyanka
Attachments
Path Model.png
Path Model.png (60.24 KiB) Viewed 7609 times
jmbecker
SmartPLS Developer
Posts: 1282
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Gender a moderator on SmartPLS 3

Post by jmbecker »

First, you need to model a direct effect from gender on mas brand image.
Second, you right click on mas brand image and choose to choose "Add Moderating Effect".
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
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:

Re: Gender a moderator on SmartPLS 3

Post by Hengkov »

Hi,

For categorical variables such as gender can not be done interactions. You must run muligroup analysis.
If there are only two values (0, 1), the interactions will be biased.

Best,
jmbecker
SmartPLS Developer
Posts: 1282
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Gender a moderator on SmartPLS 3

Post by jmbecker »

Dear Latan,
why would you think that an interaction with a 0/1 dummy variable is not allowed? It is a standard soluation that can be very meaningful.
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
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:

Re: Gender a moderator on SmartPLS 3

Post by Hengkov »

Hi Jan,

When we use a dummy variable as moderation, we will not know whether the effect of X on Y will depend on male or female (e.g, gender). This will produce different effects if run Multigroup Analysis. In addition, dummy variables can not direct interaction, because the values 0 and 1 will give the same value if multiplied.
Reference:
Hardy, M. A. (1993). Regression with Dummy Variables. Newbury Park, Sage.

Best regards,
jmbecker
SmartPLS Developer
Posts: 1282
Joined: Tue Mar 28, 2006 11:09 am
Real name and title: Dr. Jan-Michael Becker

Re: Gender a moderator on SmartPLS 3

Post by jmbecker »

Interactions with 0/1 dummies are a common practice in moderated regression. In PLS you just have to make sure that you do the correct interpretation of the effects, because of the inherent standardization of the variables. Thus, it is often easier to do a multi-group analysis in such cases. However, sometimes it is better to use the interaction approach as it has higher power, because you can use the full sample and do not have to split it into two separate samples with small sample sizes.
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
Post Reply