I Have the below residual plot as result of a multiple linear regression I have fitted.

I do not know how to interpret the results. Is it showing heteroscedasticity? There are only about 48 observations in the table and one of the predictor variables is categorical(1 and 0).

enter image description here

  • 3
    $\begingroup$ I think it is good. But there is no objective approach about this topic. so it is possible that others think it is not good. $\endgroup$
    – user158565
    Dec 2, 2018 at 5:35

1 Answer 1


Heteroscedasticity is when the variance of one variable is unequal across the range of another variable you are using to predict the first. Essentially, in the above residual v.s. fitted values plot you would expect to observe a trumpet shape. I don't personally see any.

There are tests to determine if your data follows the "equality of variance" assumption - like the levene's test - however, these can be overly sensitive, so most people like to do it by eye. There is the small hint at curvature in your residuals, suggesting that there is still some pattern left to explain in your data, which could potentially be solved by log transforming your response variable or by fitting a quadratic term to one of your covariates. If you added a pairwise matrix of each variable plotted against the others, such as one generated from pairs(df), then I might be able to give you more help.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.