I have created a linear model in R, and then plotted the residuals. If my residual plot looks like this, then which model assumptions are not appropriate?
$\begingroup$
$\endgroup$
1
-
2$\begingroup$ It looks like there is some heteroscedasticity in your residuals, also a linear fit may not be appropriate, since in the middle the residuals are mostly below the line, while at the right end they are almost all above the line. $\endgroup$– user2974951Commented Mar 9, 2020 at 10:36
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
There might be some heteroskedasticity, since for lower x values there is a much larger (vertical) spread in the residuals. Also there is maybe curvature in the plot, for x larger than about 9 all the residuals are positive, while for lower x values (around 4 to 7) most are negative. To get better advice you need to tell us more, what is the context, what is your variables, what is your modeling goal?
answered Mar 9, 2020 at 12:49
-
1$\begingroup$ I will edit my question later. I figured out what I was looking for, and will write what I did in edit $\endgroup$– user255658Commented Mar 9, 2020 at 13:12