Xametric
← Scholarship Statistics
BridgeStatistical inferenceFree

Spotting outliers with the IQR rule

IQR and the outlier rule

transport
A random sample of 40 workers living in Porirua recorded the time (minutes, door to door) of their usual morning commute. The quartiles were Q1=12Q_1 = 12 minutes, median 15 minutes and Q3=20Q_3 = 20 minutes. The shortest commute was 6 minutes and the longest was 35 minutes. Figure 1 shows the 40 commute times as a dot plot with a box plot.
Dot plot with a box plot of 40 commute times from 6 to 35 minutes; the box runs from 12 to 20 minutes with the median at 15, and one value at 35 minutes lies well to the right of the others.
Facts you may need

(a)

Calculate the IQR and the upper fence, and decide whether 35 minutes is a potential outlier.

(b)

Use the quartiles and Figure 1 to describe the shape of the distribution of commute times.

(c)

Should the 35-minute value be removed from the data? Explain.