Should I use 2 methods to detect outliers?

General
#1

Hi,

I am trying to detect outliers in my data set and I am thinking to use the standard deviation to gauge the acceptable range and potential outliers. Then, use histogram to confirm whether or not it is an outlier. Do you think this is a great method?

Thank you.

#2

Hi @Dan3

Great work there

Histogram should still get you basic visuals necessary to detect outliers.

This is all dependent on the type of dataset you are looking at, but boxplot is usually a good way to detect outliers, easy clear. If you are trying to understand two data set with similar traits, maybe one is observed and the other is predicted, you may want to consider Q-Q plot, Residual plot

1 Like
#3

Hi @Dan3 , if you make a linear model you could find some outliers with Cooks distance. But depend about you data set.

1 Like
#4

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.