MATH

profilebukysu
median.doc

Median

Noun. (Statistics.)

The median of a list of 

Example 1: The median of the numbers image1.png , image2.png , image3.png , image4.png , image5.png is image6.png. To see why this is so, arrange the numbers in order: image7.png , image8.png , image9.png , image10.png , image11.png. Then choose the middle number in the list, which is image12.png.

Example 2: The median of the numbers image13.png , image14.png , image15.png , image16.png , image17.png , image18.png is image19.png. To see why this is so, arrange the numbers in order: image20.png , image21.png , image22.png , image23.png , image24.png , image25.png. Then choose the "middle" number, which is defined to be the mean of the values in the middle of the list, that is, the mean of image26.png and image27.png.

Definition:

Given a list of image28.png numbers (not necessarily distinct), label the numbers image29.png so that image30.png.

1. If image31.png is odd, then the median of the list is the middle value in the ordered list, namely,

image32.png.

(This is to say that the median is the number in position image33.png in the list.)

2. If image34.png is even, then the median of the list is the mean of the two middle values in the ordered list, namely,

.

(This is to say that the median is the mean of the numbers in the positions just above and just below position image35.png in the list.)

Remarks:

· The median is the 50th percentile.

· Because the middle number of an ordered set of numbers (i.e., the median) will be the same whether the numbers are arranged in increasing or decreasing order, we may simply say "ordered" in the definition above without stating increasing or decreasing. However, in practice, it is more common to arrange the numbers in increasing order, as in calculating percentiles.

· The median is a measure of center of a set of numbers, like the mean and the mode. Unlike the mean, though, the median is resistant to outliers. This means that, in a sense, extreme values tend not to affect the value of median. For instance, the list of numbers image36.png , image37.png , image38.png , image39.png , image40.png is just like the list of numbers in Example 1 above, except that the extreme value image41.png in Example 1 has been replaced by the (much more extreme) value image42.png here. The median is the same in both cases, but the means are very different (image43.png in Example 1 and image44.png here.)