Measures of Central Tendency
This example shows how to compute and compare measures of location for sample data that contains one oulier.
Generate sample data that contains one outlier.
x = [ones(1,6),100]
x = 1 1 1 1 1 1 100
Compute the geometric mean, harmonic mean, mean, median, and trimmed mean for the sample data.
locate = [geomean(x) harmmean(x) mean(x) median(x)... trimmean(x,25)]
locate = 1.9307 1.1647 15.1429 1.0000 1.0000
The mean (mean) is far from any data value because of the influence of the outlier. The geometric mean (geomean) and the harmonic mean (harmmean) are influenced by the outlier, but not as significantly. The median (median) and trimmed mean (trimmean) ignore the outlier value and describe the location of the rest of the data values.