Assume I have the following regression model:
$Y=\beta_0 + \beta_1*X_1 + \beta_2*X_2 + \beta_3*X_3 + \epsilon$
If I run the same model without introducing the variable $X_3$, can I interpret the new residuals as the values of $Y$ minus the effects of covariates $X_1$ and $X_2$ (but the effect of $X_3$ on $Y$ remains)?