Confused which test to use

[Mirte Roos]

Good day everyone,

Currently I am trying to test a hypothesis which is strucutred the following:
- Independent variable (yes/no, having a team)
- dependent variable (continuous, success rate)
- moderator (3 options, a layer present, 2 layers present, 3 layers present).

Do I need to use a one-way Anova now, or what would you recommend me?
The moderator differs depending on hypothesis. However, they are always categorical and not a ranking or something similar.

Thank you!!

[Carlo Lazzaro]

Mirte:
there's (almost) nothing that -regress- can't do even better than anova.
That said, you can try:

regress <depvar> i.team i.moderator

[Joseph Coveney]

Observational study jargon remains arcane to me, but doesn't what people call a "moderator" involve interaction?

fracreg logit success_rate i.has_team##i.layer_tally
contrast r.has_team#p.layer_tally

[Marcos Almeida]

If by ‘moderator’ you meant an endogenous covariate model, you may take a look at the ERMs (extended regression models).