ancom model equation

[DennisBeest]

Looking at your ancom code (and the example from the help file) I noticed you set up your model equation as this:

tformula = formula(paste("x ~", main_var, "+", adj_formula), sep = " ")

which means the main variable is in first place in the equation. Later in the code you do an anova with this. The default with anova(lm(..)) is a type 1 anova, which means the variables are added sequentially.

What this means is that in the currrent setup the SS/p-value you get for main_var is not adjusted for adj_formula.

[Maggie8888]

Please be more specific for your question. Are you wondering the p-value for main effect is not considering adjustment for other covariables in the model? And which function are you referring to?