IQR is a vital concept, particularly for visualizing data spread and detecting unusual values.
IQR measures the statistical dispersion that describes the middle 50% of your data.
Q1 - Value below 25% of data.
Q2 - Midpoint - 50% of scores are below this value
Q3 - Value below which 75% of the data falls
Maximum Score
IQR = Q3-Q1
Example walkthrough:
[0,13,14,19,21,23,40]
7 values: Q2 = 7+1 / 2 = 4th positon
Now 3 on each side.
3+1 /2 = 2, 2nd value, so now you can say
Q1 = 13 Q2 = 19 Q3 = 23
IQR = 23-13 = 10
Lower Bound = 13 - (1.5 * 10) = -2 Upper Bound = 23- (1.5 * 10) = 38
**NOTE - the 1.5 is specific to the boxplot method.
lost? ~ Index - Data Engineering