I suppose it could just be preference, or it could be that you want to use the method that you know (I first learned about the IQR method before any other method, so I tend to stick with that). In certain situations, you may find that your data is more visually appealing with certain methods. Thinking about it now, we could reduce some of the outlier clutter by modifying the chart to include everything from 2% to 98% instead of using the IQR to find outliers

Did this answer your question?