In G Power - which should I use for a hierarchical regression with continuous variables? t test - linear multiple regression: fixed model, single regression coefficient OR OR f test - linear multiple regression: fixed model, R2 deviation from zero. What is the difference?
Hello Bonnie,
The answer depends on which step in the hierarchy you're trying to evaluate.
1. If your base model includes only the target variable(s) (as in an "unadjusted," "raw," or "crude" model), then use: Fixed model, R2 deviation from zero.
2. If your base model includes only "control" variables, then again, use Fixed model, R2 deviation from zero.
3. If your second-tier model is intended to look at the added explanatory power of one or more target variables, given the "control" or previously entered variables, then use Fixed model, R2 increase.
Good luck with your work.
- Badges
- Science topic