How does JMP derive 95 percentile of a data set with limited data points?

[#1]

For a data set with 16 points shown below,

ID	Data
1	-2.53802
2	-0.98877
3	-0.59679
4	-0.45444
5	-0.26573
6	-0.2027
7	-0.10798
8	0.124422
9	0.156302
10	0.162865
11	0.469599
12	0.667054
13	0.787399
14	1.209767
15	1.414172
16	1.718419

Tabulate the data generates the following stats:

Min	-2.53802
Mean	0.097223
Quantiles50	0.140362
Quantiles95	1.718419
Max	1.718419
Std Dev	1.023793

I am curious to know how JMP derives 95 percentile or for any random percentile. 

Thanks!

[#2 · Accepted solution]

Here is the information you want, taken from

Help==>Books==>Basic Analysis

quantile.png