I need to compute three two-way ANCOVAs for my 2×2 between subjects design.

My dependent variable is: Extraversion Rating

My independent Variables are: Pictures (high or low, 1 or 0) x Followers (high or low, 1 or 0)

My three covariates are: age, gender and rating of frequency of Instagram use (rated from 1 to 5)

Now I´m confused because: first of all (except for age as a continuous variable which i z-standardized) can i even use gender or the rating scale as covariates in an ANCOVA? i tried looking it up but found little information in my textbooks

Furthermore, probably related to this, my linearity assumptions are not met (ggscatter). For the rating they look like this:

age looks like this:

For age I tried log/sqrt/inverse transformation (none of them helped). a quadratic term/polynomial transformation of age did not really help either.

Is use of WRS2 package enough?

Do i just need to use a different method than ANCOVA?