Measures of Location: Computing Percentile
soleadea
5 years 3 months 2 weeks
Thanks to this short post, you get a useful piece of advice related to measures of location, recall an important formula and are offered a practical tip.
Remember!
Before computing any measure of location (like percentile, median, etc.), sort observations in an ascending or a descending order.
Formula for the location of a percentile
\(L_y=(n+1)\times\frac{y}{100}\)
- Where:
- \(y\) – percentile,
- \(L_y\) – location of the percentile,
- \(n\) – number of observations.
Measures of location: practical tip
If the location of the percentile is not an integer, we need to apply linear interpolation to compute this percentile. Have a look at the example.
Example:
There is a set of numbers given in the ascending order: 2, 12, 23, 34, 45, 55, 65, 83, 87 and 89. Determine the value of the 20th percentile.
Solution:
\(n=10\)
\(L_{20}=(n+1)\times\frac{20}{100}=(10+1)\times\frac{20}{100}=2.2\)
\(P_{20}=X_2+(2.2-2)\times(X_3-X_2)=\\=12+0.2\times(23-12)=12+0.2\times11=14.2\)