OddThinking

A lecturer in statistics once explained to me that when you had a large sample of measurements, it is sometimes necessary to identify the “outliers”; values which are either very suspicious (if there is a risk of experimental error) or very interesting because they are so far outside of the normal range of values.

There’s a quick classification of outliers based on how far away they are from the mean value – either they are classed as “near” or “far” – for those who have to know, it depends on how many IQRs (Interquartile Ranges) they are from the mean value.

I didn’t get a sudden deep understanding of statistics that day.

I didn’t get a mathematical tool that I have found useful again and again.

I did, however, get a sudden new goal in life. One day, I would find the opportunity to call a statistician a “far outlier”.