New method uncovers hidden outliers in data, revolutionizing data analysis.
Detecting outliers in a group of data points can be tricky, especially when there are many outliers present. Traditional methods often miss these outliers because they are based on average values that can be skewed by the outliers themselves. To address this issue, the researchers suggest using more robust calculations to identify outliers accurately. In cases of regression data, the usual approach can also hide outliers, but a more robust method can reveal them. The researchers also introduce a new way to visualize the data, which helps classify observations as normal, outliers, or leverage points.