Age and Gender

[Isabelle Deschamps]

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

[Carlo Lazzaro]

Isabelle:
welcome to this forum.
Despite being symphatetic with beginners (as I'm still one of these in some respects), you should have carefully read the FAQ before your first post.
As you will see, attachments are not encouraged and spreadsheets will not be downloaded because they are far from Stata format and, especially, because of possible active contents.
That said, it is not absolutely clear what you're after. Please note that the FAQ reccommend to share what you typed and what Stata gave you back and/or to post an example/excerpt of your data via -dataex-
As an aside, please do not consider this forum as a place where others can/will do research on your behalf or devote their time to decypher what you have in mind but forgot to explain: we will certainly let you down.

[Bruce Weaver]

Hello Isabelle. As Carlo has noted, the FAQ gives some excellent advice about how to write questions that are more likely to prompt helpful responses. Meanwhile, I'll take a stab at clarifying things a bit. Here's what I got from your first post:

Outcome variable: Entry Intention (1=Yes, 0=No) <-- Is this right? If it is binary, you probably want a logit or probit model. I prefer logit, as it yields odds ratios.

Explanatory variables

Gender: male (116), female (97), non-binary (27)
Age category: 18-24 (195), 25-34 (27), 35-44 (7), 45-60+ (11)
Type of Window Display: 4 types

Q. Is Type of Window Display a repeated measures variable? I.e., is each subject exposed to all 4 types (or at least to more than one type)? If so, you need to use a method that takes into account the correlated nature of repeated measures data (e.g., GEE or multilevel model with repeated observations clustered within subjects).

You said: "And I want to know if age has an impact on the attraction towards technology-based window."

It sounds like you want to include the Age x Display Type interaction in your model.

HTH.