Assumption of linearity

When gender is removed from the regression the scatter plot of residuals versus predicted values changes, does this mean gender is why the plot from my original regression is just two vertical lines? Is it showing the male and female scores separately? The first photo is with gender included and the second is without. I am trying to deal with establishing if my data meets the assumption of linearity so if anyone can provide any advice that would be much appreciated.



Ambassador to the humans
Linearity isn't the issue. But it looks like whatever the continuous predictor you're using is doesn't explain much. The R² is tiny. In your original plot it isn't two completely vertical lines - it just looks like that because the other variable besides gender barely impacts the prediction. What are the coefficients for your models.