Bonjour,

First and foremost, I have to warn you that Im such a beginner on STATA & statistics overall ...
But with regards to my master thesis on fashion windows displays & consumer behavior, I had to do statistics analyses !
Anyway, Im facing now the issue of including gender & age, as covariates/moderator variables of the model.

gender: male (116), female (97), non-binary (27)
age: 18-24 (195), 25-34 (27), 35-44 (7), 45-60+ (11)
In total 240 participants, but the sample sizes are different! I think it is a problem.

I tried to code both as dummy variables, but received the message "no observations".

I want to study if including genders identities influence Entry Intention (Y outcome variable) regarding f4 different Fashion Windows Displays (independent variables: X1Mannequins X2PunchyWarmColors X3InteractiveTechnology X4CreativeDesign)
And I want to know if age has an impact on the attraction towards technology-based window.

- maybe: dummy variables are not the solution?
- others models should be used like: a binary response model?
- maybe ANOVA? by separating the population using a median split?

I already reg per gender & the three models are significant p-value <0,05, R²<1.
Also, F-Test is OK.
VIF: checked as well.

I hope that I have expressed myself correctly.
I include here my Excel file.

Brgrds,
Isabelle Deschamps