Multi-level modeling

[Matthew Fuller-Tyszkiewicz]

Hello list.

 

I wish to conduct multilevel regression modelling to evaluate how the impact of cultural messages upon participants’ beauty ideals differ by country. [note that cultural messages is measured as 5 separate variables - 1 each for five different types of messages] I anticipate that the strength of the relationship between cultural messages and beauty ideals will differ  by country, and also by gender. However, I am unsure about how to include gender in my analysis. Are any of the following options suitable:

 

1)      Conducting separate multi-level regression models for males and females

2)      Conducting one multi-level regression model using gender as a potential moderator of the relationship between cultural messages and beauty ideals?

3)      Conducting one multi-level regression model using gender and country as variables at level 2?

 

Thanks in advance for your assistance,

Matt.

 

 

[Bruce Weaver]

Matthew Fuller-Tyszkiewicz wrote

Hello list.

I wish to conduct multilevel regression modelling to evaluate how the impact of cultural messages upon participants' beauty ideals differ by country. [note that cultural messages is measured as 5 separate variables - 1 each for five different types of messages] I anticipate that the strength of the relationship between cultural messages and beauty ideals will differ  by country, and also by gender. However, I am unsure about how to include gender in my analysis. Are any of the following options suitable:


1)      Conducting separate multi-level regression models for males and females

2)      Conducting one multi-level regression model using gender as a potential moderator of the relationship between cultural messages and beauty ideals?

3)      Conducting one multi-level regression model using gender and country as variables at level 2?

Thanks in advance for your assistance,
Matt.

Run a single model with gender as a factor, and with interaction terms involving gender if they improve the fit.  But regarding point 3 above, if gender is a level 2 variable, I assume each subject is exposed to all 5 cultural messages.  Is that right?  Also, if gender is a level 2 variable, isn't country a level 3 variable?  I.e., if I understand you, messages are nested within people (males & females), who are nested within countries.