lisakki2501
Joined: 05 Feb 2009
Posts: 1
|
 Posted: February 5, 2009 6:05 PM Post subject: Interquartile Range |
 |
|
When I enter in the following set of data into my TI-89:
28, 42, 58, 48, 45, 55, 60, 49, 50
And I type in 1 Var Stat, it gives me
Q1X=43.5
MedX=49
Q2X=56.5
Now, the formula to calculate the index is i=(p/100)*n , where p is the percentile you want to calculate. If i is an integer, Q1 would be the average between the i and i+1 positions. If not an integer, you simply round up and use that as the position.
So for the first quartile, i=(25/100)*9=2.25, which rounds up to 3. After sorting the list, the 3rd number on the list happens to be 45. Therefore, Q1=45. After double checking the data on excel, I get the same result.
My question is, why is the TI 89 wrong about the quartiles? The only thing it got right out of those 3 is the median. If you think about it, there is no logical way that Q1 can not be an integer in a set of 9 numbers if all the numbers are integers themselves. |
|
