The case with the largest negative difference between the observed value and the predicted value on the outcome. Depending on the research question, there might be a specific interest in the case for which the model performs worst and yields a larger predicted value.
Arguments
- lmobject
Object generated with
lm
Examples
df <- lm(mpg ~ disp + wt, data = mtcars)
most_overpredicted(df)
#> Ferrari Dino
#> -3.40868