You need to get the actual value that corresponds to the particular percentile.
So basically you need to find P(X <= x) = a for some x given a value of a (20th percentile implies a = 0.2).
Lets say I have a list of values 1 2 3 4 5 6 7 8 9 10.
and I want to find the 20percentile I whould do this.
20/100*10=2 position two which just happens to have the value of 2 thats fine.
Now lets say I have this list of values 1 2 3 4 5 6 7 8 9
and I want to find the 20 percentile I whould do this.
20/100*9=1.8 because this now produces a decimal do just round up to position two
which just happens to have the value of 2.
because if I find the average of the first ad secund value I get a decimal again ie 1.5.
Conclution: If theres an even number of values just take that value and if theres an odd
number of values round up and take that value?
First you rank the observations in ascending (increasing) order and find the probability distribution (from the frequency table).
Once you have done this then you find the x such that P(X <= x) = a where a is your percentile.
This is the general method for finding percentiles.